|
Conference Papers | 2005 Victorian Conference Papers
NITROGEN
REMOVAL USING TERTIARY FILTRATION
Suzy
Hatch, Production
Officer,
Sydney Water
Colum Kearney, Production
officer, Sydney
Water
 DOWNLOAD
(pdf
102 KB)
BACKGROUND
In 2002 both West Hornsby and Hornsby Heights Sewerage
Treatment Plants (STP) completed a Biological Nutrient
Reduction (BNR) upgrade. Prior to the upgrade the plants
were achieving Total Nitrogen (TN) of approximately
25mg/l. The aim of the upgrade was for both plants to
achieve a TN of 5mg/l as a 90 percentile in the effluent.
Realistically that means the plants have to target a
TN of 3mg/l to achieve compliance. Included in the BNR
upgrade was the installation of a methanol dosing facility
providing an external carbon food source for more removal
of TN.
Due
to the nature of BNR, process designers were unsure
how to combine and quantify the internal organic carbon
food source from the fermenter with the methanol in
order to achieve desired TN levels. It was up to the
production team to test various quantities and dosing
points of the Methanol to optimise the best results
for nitrogen removal from the wastewater. Both plants
were commissioned in June 2002 and after 18 months of
process changes the plants were still only achieving
approximately 7mg/l of TN in the plant effluent. We
were a long way from reaching the goal of the upgrade.
At a similar time a pilot project of dosing methanol
to the filters was being carried out at Quakers Hill
STP. Although the trail showed some promising results
for TN reduction, it had to be abandoned due to the
nature of their shallow bed sand filters, which were
unable to cope with the added solid production. It was
then recommended that further testing be carried out
at the Hornsby Plants to see what affects this dosing
would have on Dual Media Filters.
In November 2003 both Hornsby plants conducted trials
of dosing Methanol to the tertiary filters to determine
the ability of the filters to further reduce Nitrogen
prior to disinfection.
1.0
PROCESS DESCRIPTION - HORNSBY TREATMENT PLANTS
The
Hornsby Treatment plants both operate under a 5 stage
modified Bardenpho process which is illustrated in the
diagram below.
Figure
1: 5 Stage Bardenpho Biological Nutrient Removal Process-
Hornsby Heights STP & West Hornsby STP.
Primary
effluent, fermented sludge and Return Activated Sludge
(RAS) mix in the Anaerobic Zone which provides
an environment free of dissolved oxygen where much of
the biological phosphorus release occurs. The flow then
enters a three way splitter where Mixed Liquor Recycle
(MLR) is returned to the process and flows into the
Anoxic Zone. This zone provides an environment
free of dissolved oxygen where denitrification occurs.
Flow then enters the Aerobic Zone where the presence
of dissolved free oxygen, ammonia and carbonaceous components
are biologically oxidised and phosphorous uptake occurs.
Within the Secondary Anoxic Zone methanol or
ethanol is added to provide an external carbon source
for denitrification. Flow then enters the Re-aeration
Zone which provides a small amount of additional
nitrification and phosphorus uptake before the secondary
clarifiers.
1.1 Tertiary Filters at the
Hornsby Treatment Plants
Both Hornsby Heights & West Hornsby have Dual Media
filters containing a top layer of Anthracite and lower
layers of sand and fine gravels. Figure 2 below illustrates
the design process of a Dual Media Filter.
Figure 2: Dual Media Tertiary
Filtration Process Diagram (West Hornsby UPG).
The
secondary effluent flows from the filter inlet chamber
through the filter media. Suspended solids in the secondary
effluent are trapped in the filter bed. Effluent then
flows through the underdrain to the holding tank then
flows to UV disinfection prior to discharge. The filters
are regularly backwashed to remove any trapped suspended
solids. The suspended solids then drain to the foul
water tank and are pumped back to the head of the plant
for treatment.
1.2 What is Nitrogen Removal?
Nitrogen removal within wastewater treatment can best
be described as a combination of two biological reactions
notably nitrification and denitrification. This two
step process is achieved by a variety of bacteria that
function under different environmental conditions in
separate zones within the wastewater treatment facility.
»
NITRIFICATION:
NH3+
O2+ alkalinity NO3+ newcells +H2O
Figure 3: Schematic diagram
of the nitrification process. (Hornsby Heights BNR UPG).
Nitrification occurs within the aerobic zone (environment
where free oxygen is present). Micro-organisms in activated
sludge known as autotrophs (nitrifying bacteria) require
this oxygen and alkalinity within the aerobic zone to
convert the ammonia into nitrites and then nitrates.
»
DENITRIFICATION:
NO3 + Organic matters N2 + CO2 + H2O
Figure 4: Schematic diagram
of the Denitrification process. (Hornsby Heights BNR
UPG).
Denitrification
occurs in the zone were no free oxygen is available
- the anoxic zone. Another group of bacteria known as
heterotrophs convert the nitrite/nitrate produced during
nitrification to nitrogen gas. As can be seen from figure
4, these bacteria also require a carbon source (food)
for this reaction to occur. This carbon food source
required for denitrification can be classified as organic
carbon i.e. - occurring naturally in sewage or fermented
sludge. The other is an external carbon source. At the
Hornsby treatment plants this external source of carbon
is methanol or ethanol.
The
main reason why Methanol or ethanol is a good carbon
source is the nature of their chemical makeup. To quote
from the US Methanol Institute, "Methanol serves as
a carbon source for bacterial bugs. Accelerated by the
addition of methanol, anaerobic bacteria convert the
nitrate to harmless nitrogen gas, which is vented into
the atmosphere".
2.0
HORNSBY HEIGHTS STP
2.1 Procedure
| 1 |
To
facilitate the trial minor changes were required
to the Methanol dosing system |
| 2 |
Baseline
results were determined by analysing the filter
effluent for nitrate (NOX) and ammonia over a period
of 7 days with no changes to the system. The plant
lab is not able to analyse for TN so nitrate level's
were used. |
| 3 |
Commenced
with a dose rate of 5 mg/L of methanol to filters
flow paced to the primary effluent flow, increasing
to 10 mg/L overtime. |
| 4 |
The
time between backwashes of the filters for most
of the trial was 18 hours. As the filters were starting
to fail the time between backwashes was decreased
to 16 hours and then again to 12 hours before ceasing
the methanol dose. |
| 5 |
The
parameters kept constant to avoid results being
inconclusive were;
-
30 mg/L of methanol dosed to the secondary anoxic
-
Fermenter sludge age
-
dissolved oxygen levels
- Alum,
SPL and lime dosage rates
- Mixed
liquor recycle rate
- %
availability of equipment
- Samplers
set up to take 4 hourly samples from clarifier
effluent and plant effluent (pre and post the
filters)
|
2.2
Results
| 1 |
At
a dose rate of 5mg/L of Methanol to the filters
the NOX results varied showing some decrease but
nothing substantial. |
| 2 |
When
the dose rate was increased to 10mg/L an average
decrease across the filters was 1.3mg/L |
| 3 |
Due
to the solids increase in the filters the trial
had to be abandoned. We were alerted to the problem
as the filters were continually backwashing on head
loss. We realised that two out of the four filters
had not been upgraded for ten years and required
sand and media replacement. |
Figure 5: Nitrate results for
the CE & FE during the Methanol trial at Hornsby Heights
STP. (Note: period from 10/11/03 to 19/11/03 problems
with CE sampler)
3.0 WEST HORNSBY STP.
3.1 Procedure
| 1 |
Baseline
results were determined by analysing the filter
effluent for nitrate (NOX) and ammonia over a period
of 7 days with no changes to the system. The plant
lab is not able to analyse for TN so nitrate level's
were used. |
| 2 |
All dosing to other points was stopped. |
| 3 |
On
the 8th of November 2003 the dosing of methanol
to the Tertiary Filters began at 5mg/L flow paced
to the primary effluent flow. |
| 4 |
After
7 days the dose rate was increased to 10mg/L for
a further 4 weeks. |
| 5 |
On the 4th of December 2003 the dosing rate was
further increased to 15mg/L. |
| 6 |
On
the 17th of December West Hornsby STP received a
delivery of ethanol; therefore, the product was
a mixture of both methanol and ethanol. |
| 7 |
22nd
of December dosing restarted to the Secondary Anoxic
at 15 mg/L. |
| 8 |
During
the dosing of Methanol / Ethanol to the Tertiary
Filters at West Hornsby STP the backwash time remained
at 24 Hours. This was to allow as much biomass growth
on the filter bed increasing biological activity
thus gaining as much nitrogen removal as possible
from the filters within the backwash run time. |
| 9 |
It
must be noted that as well as testing of Nitrates
other analysis on Ammonia and COD were carried out.
There was no significant variation in these results
over the period in which the trial was carried out.
|
| 10 |
Early March 04, dosing to the filters was stopped
as the filters became blocked and continually backwashing
on head loss. |
3.2
Results
| 1 |
It was not until the dose rate was increased to
15mg/L that the trial began to show a clear difference
between the amount of nitrate recorded in the Clarifier
Effluent (CE) compared to the Final Effluent (FE).
|
| 2 |
Approximately
one week after dosing was restarted to the secondary
anoxic there was a significant reduction in the
overall nitrate levels in the plant effluent. |
| 3 |
The
average difference between the nitrate present in
CE and that in FE for the period from the 08/12/03
to the 19/01/04 was 2.52mg/L. |
Figure 6: Nitrate results for the CE & FE during the
Methanol/Ethanol trial at West Hornsby STP.
Figure 7: Difference between Nitrate present in CE and
FE during the methanol/ethanol trial.
Figure 8: EPA Licence results (TN) from the 19/11/03
to the 30/03/04, when methanol/ethanol was being dosed
to the Tertiary filters at West Hornsby STP.
4.0
CONCLUSION
As can be seen from the results at the Hornsby treatment
plants the ability to dose methanol or ethanol to the
tertiary filters can reduce the nitrogen levels in plant
effluent. However, the degree of nitrogen removal is
vulnerable to many factors specific to each individual
treatment plant or process condition.
Probably the most significant factor that impacts on
the nitrogen removal process in the tertiary filters
is the capacity and backwash period of the filters.
A problem highlighted at Hornsby Heights STP was that
two of its four filters operated poorly and blocked
up very quickly with the addition of methanol. Further
analysis of these two filters showed that they had approximately
40% less media than the other two filters, consequently
these two filters are to be refurbished.
West Hornsby STP with eight slightly larger filters
running on a 24-hour backwash time proved that it could
remove further nitrogen from the process. However, after
a period of time the filters here started to block up
and could not function properly. It is widely accepted
that perhaps the best way to alleviate the issue of
filters blocking up due to biomass growth and increased
levels of nitrogen gas accumulation is what is termed
a 'bumping procedure' at the filters. This involves
aerating the filter media occasionally between backwashes
so as to relieve nitrogen gas accumulation and extend
the filter run time, thus maximising the filters ability
to remove more nitrogen from the process.
This
bumping procedure was not carried out at either of the
Hornsby treatment Plants as it was uncertain what, if
any long-term detrimental effects this may have on a
dual media filter. Researching the Internet it was found
that several treatment plants in Europe and the US have
installed filters (BIOSTYR & BIOFOR) that are specifically
configured for nitrogen removal. Methanol addition to
the unaerated sections of these filters is suitable
for denitrification.
This
trial has proven that it is possible to remove nitrogen
using tertiary filters. At West Hornsby STP this rate
of removal was on average 2.52mg/L between the 8th of
December 03 and the 19th of January 04. However, it
is recommended further studies be carried out to assess
if the additional costs involved in dosing methanol/ethanol
to the filters and higher energy consumption with increased
backwashing or bumping of the filters is economically
viable in terms of process results and plant operations
in the long-term.
5.0
REFERENCES
| 1 |
Daigger
GT, Biological Nutrient Removal. Colorado |
| 2 |
Koch
G & Siegrist H, (1997) Denitrification With Methanol
In Tertiary Filtration at A Wastewater Treatment
Plant- Zurich- Werdholzli. |
| 3 |
Methanol Institute USA, Wastewater Treatment
With Methanol Denitrification- WWW.methanol.org |
| 4 |
Solley David, (2002) Upgrading of Large wastewater
Treatment Plant for Nutrient Removal. Churchill
Fellowship Report. |
| 5 |
West
Hornsby/Hornsby Heights- Unit Process Guideline
(UPG), BNR Process. |
| 6 |
West
Hornsby/Hornsby Heights- Unit Process Guideline
(UPG), Filtration Process. |
DOWNLOAD
(pdf
102 KB)
|
