STARTUP AND OPERATION OF A BIOLOGICAL

AERATED FILTER (BAF) FOR SIMULTANEOUS BOD

REMOVAL AND NITRIFICATION AT THE LOU ROMANO

WATER RECLAMATION PLANT

WINDSOR ONTARIO CANADA

International Joint Commission

“Water & Wastewater Treatment Best Management Practices Forum”

Monday, March 26, 2012

Wayne State University, Detroit MI

Paul Drca

Manager Environmental Quality

City of Windsor

Presentation Outline

1. Introduction

2. The Lou Romano Water Reclamation Plant (LRWRP)

3. Pilot Plant Studies

4. Selection of Biological Aerated Filter (BAF) for Secondary

Treatment

5. LRWRP MOE Certificate of Approval

6. BAF Startup & Challenges Encountered

7. LRWRP Effluent Quality

8. Conclusions & Questions

Introduction

Prior to 2007, the City of Windsor’s Lou Romano

Water Reclamation Plant, formerly called the West

Windsor Pollution Control Plant, was the largest and

one of the last remaining primary chemical physical

wastewater treatment plants in the Great Lakes

basin. The plant effluent is directly discharged to the

Detroit River, one of 43 areas of concern identified by

the International Joint Commission (IJC) requiring

remedial action with respect to beneficial use

impairments.

Introduction

Realizing the need to upgrade the LRWRP to

secondary treatment, the City performed pilot plant

studies in 1989-91 and 1994-96 to investigate

innovative alternatives to conventional secondary

treatment in an effort to determine the preferred

process for their needs and to minimize wastewater

treatment plant expansion costs

Environment Canada’s Great Lakes Sustainability

Fund (GLSF) provided funding assistance for the pilot

plant studies

The LRWRP - Background

• Construction of the plant began in 1967 and it

commenced operation as a 110,000 m3/d primary

treatment plant in 1970

• In 1974 it was converted to a physical-chemical

treatment plant incorporating phosphorus removal

• Numerous upgrades and expansions of the plant

occurred throughout the 1980s

The LRWRP - Upgrade &

Expansion

• In 2005, the City commenced with an upgrade and

expansion of the LRWRP that included the expansion

of primary treatment rated capacity from 163,700

m3/d to 272,800 m3/d and the construction of a BAF

for secondary treatment with a rated capacity of

218,000 m3/d.

• The plant is designed to handle a peak flow of two

times the rated capacity.

The LRWRP - Upgrade &

Expansion Other Work included:

• New coarse bar screens and conveyors

• New primary sludge pumphouses

• Primary effluent pumping station

• Ultraviolet (UV) disinfection (secondary effluent)

• Sodium hypochlorite (bypass flows) disinfection

facility

• New centrifuge sludge dewatering equipment

The $110,000,000 upgrade and expansion of the

LRWRP is the largest single project ever undertaken

by the City of Windsor

The LRWRP BAF Construction

Construction of Secondary Treatment

Biological Aerated Filters

Pilot Plant Studies

• Pilot testing was completed in two phases

• 1989-91 - Rotating Biological Contactor (RBC),

Trickling Filter/Solids Contactor (TF/SC), Biological

Aerated Filter (BAF) and Modified Activated Sludge

processes were piloted

• 1994-96 - TF/SC and BAF were selected for further

testing as to their suitability for the LRWRP upgrade

• Effluent objectives during the study were 15, 15, 0.5

and 3.0 mg/l BOD, TSS, TP and NH3-N respectively

Biological Aerated

Filter Pilot Plant

Selection of Biological Aerated Filter

(BAF) for Secondary Treatment

• BAF was selected as the solution for the City’s need

to upgrade the LRWRP to secondary treatment due

to its compact footprint and ease of future

expandability on the site.

• The advantages of the BAF are both financial and

environmental, capital costs are approximately 25%

less than conventional alternatives and the BAF

footprint is approximately one third of what a

conventional secondary treatment plant of the same

capacity would occupy.

Biological Aerated Filter Process

LRWRP MOE Certificate of

Approval

Parameter Effluent Limit *

CBOD5 15.0

TSS 15.0

TP 0.5

Un-ionized NH3 0.1

Acute Lethality to Rainbow

Trout & Daphnia Magna

Non acutely lethal

E. Coli 200 organisms/100ml

pH 6.0 – 9.5

*Effluent limits are monthly average concentrations in mg/l

unless otherwise indicated

LRWRP MOE Certificate of

Approval

COA Compounds

Lindane

Chlordane

Aldrin/Dieldrin

Cadmium

Mercury

PCBs

Benzo(a)pyrene

Hexachlorobenzene

COA Compounds –

plant effluent analyzed

quarterly

LRWRP MOE Certificate of

Approval - Compliance

• As of August 2010, the LRWRP has consistently met

C of A limits

• A result of proactive process improvements including

basing BAF cell run time on loading criteria and

primary effluent quality

• No COA compounds other than trace amounts of

mercury and cadmium detected in effluent

• Plant effluent has consistently been non acutely

lethal to Daphnia Magna and Rainbow Trout

• UV system has successfully replaced chlorine for

disinfection

BAF Startup & Challenges Encountered

• 80% BOD removal within 3 weeks of startup

• Steady nitrification by June 2008

Challenges

• Foaming

• Soluble phosphorous deficiency

• TSS compliance

• BAF cell loading

• Media loss

• Primary effluent & backwash supply tank screening

• Backwash cycle time

0

20

40

60

80

100

120

140

160

180

200

BO

D (

mg

/l)

BOD Removal Within 3 Weeks of Startup

INF BOD PRI EFF BOD BAF EFF BOD

Foaming

• Lots of foam covered BAF cells on startup

• Primary effluent dosed with defoamer and cells were

covered with tarps to control foam

• Upon establishment of sufficient biomass, foam

subsided

• Foaming is still an issue at times

• Tarps have been removed, defoamer is still required

at lower doses

Soluble Phosphorous

Deficiency

• Pilot plant testing indicated PE was soluble P

deficient

• Full scale plant was designed with H3PO4

supplementation

• H3PO4 dosed to deliver 1 mgl-1 P to PE to maintain a

concentration of 0.2 - 0.3 P in BAF effluent

• P supplement is expensive – ongoing project with

University of Windsor to optimize P removal in PE to

eliminate or minimize need for P supplement

TSS Compliance

• On startup, LRWRP did not consistently meet TSS

limit of 15 mgl-1 on a daily basis

• Variable TSS loading from PE was found to be a

contributing factor due to dewatering centrate, BAF

backwash solids return to primary clarifiers, and wet

weather flows.

• Since the summer of 2010, BAF cells have been

backwashed based on solids loading instead of fixed

filtration time.

TSS Compliance Solution

• Monthly strong backwashing frequency was doubled

and spread out through the month

• Backwash procedure was revised for efficiency

• Monitoring and maintaining primary clarifier surface

overflow rate (SOR) - target of 20 m/h SOR

• An additional chemical feed line for alum was

extended to the backwash waste discharge pipe (just

prior to primary treatment)

• Cell cycle time is set based on BAF media loading

criteria and primary effluent quality

Data Constants

TSS 1.26 30 day maximum loading in - kg TSS/m3d 

TKN 0.59 31 day maximum loading in - kg TKN/m3d 

Cell Area 144 m2

Media Depth 3.9 m

Media Volume 561.6 m3

Hours/day 24

Avg. Daily

Flow

(MLD/day)

Avg.# of

Cells in

Filtration

Primary

Effluent

TSS

(mgl-1)

Primary

Effluent

TKN

(mgl-1)

Max

Duration

of TSS

Cycle

(h)

Max

Duration

of TKN

Cycle

(h)

Max

Filtration

Time

(h, min of

TSS/TKN)

Filtration

Time

(h, 3 day

running

avg)

147.0 8.0 39.0 10.0 23.7 43.3 23.7 24

125.0 6.8 39.0 10.0 23.7 43.3 23.7 24

132.7 7.2 37.0 20.0 24.9 21.6 21.6 23

173.6 8.5 60.0 20.0 13.9 19.5 13.9 20

107.5 6.8 60.0 20.0 17.9 25.2 17.9 18

124.5 6.9 60.0 34.0 15.7 13.0 13.0 15

125.0 7.0 10.0 34.0 95.1 13.1 13.1 15

2010 LRWRP BAF TSS

0

20

40

60

80

100

120

01/01/2010 01/04/2010 01/07/2010 01/10/2010 01/01/2011

TS

S (

mg

/L)

BAF Influent

BAF Effluent

• Changed from filtration time to solid loading in the summer of 2010

• Significant improvements in LRWRP effluent quality

Media Loss

• Media loss was excessive during 2008-09

• Geyser type eruptions of media during backwashing

resulted in media being returned with backwash

wastewater

• Media was recovered from the backwash waste tank

• Media loss appears to have subsided since process

improvements implemented as of Aug 2010

• Evaluation of media loss is ongoing

Media loss experienced during backwash

Backwash Cycle Length

• The length of time to complete a cell backwash is a

concern

• This is only an issue when backwashes are queued

up

• Using the BAF configuration to full advantage, it is

possible to start a backwash before an ongoing

backwash is completed through programming

modifications

• These modifications will result in up to a 45 minute

decrease in average backwash time

Primary Effluent & Backwash

Supply Tank Screening

• Nozzles delivering PE and backwash rinse water to

BAF cells require protection from plugging

• 4 channels fitted with fine curved screens filter PE to

the BAF cells

• 4 screens filter BAF effluent prior to storage in the

backwash supply tank

• Weekly power washing of fine curved screens is

required

• Backwash supply tank screens are washed as

required – up to a 3 day interval between service

Primary Effluent Fine Curved

Screens

Backwash Supply Tank

Screens

Future Work

• Media loss quantification

• Media fluidization criteria

• Primary clarifier optimization

• Backwash efficiencies

• Energy management

LRWRP Improvements in

Effluent Quality

LRWRP Effluent

2006 vs 2010

Concentration (mg/l)

Parameter 2006 2010*

TSS 25.7 5.1

TP 0.58 0.41

BOD** 46.8 2.8

TKN 16.2 3.7

NH3 10.6 1.8

NO2-NO3 0.9 5.6

*Based on average daily concentration Aug - Dec 2010

**2006 BOD is TBOD, 2010 BOD is CBOD

LRWRP Improvements in

Effluent Quality

LRWRP Effluent

2006 vs 2010

% Removal Efficiency

Parameter 2006 2010*

TSS 81.7 97.3

TP 84.5 91.3

BOD** 63.1 98.3

TKN 19.4 85.6

NH3 6.5 87.1 *Based on average daily concentration Aug - Dec 2010

**2006 BOD is TBOD, 2010 BOD uses TBOD for influent, CBOD for effluent (as per C of A)

LRWRP Improvements in

Effluent Quality

LRWRP Effluent

2006 vs 2010

Annual Loading (tonnes)

Parameter 2006 2010* Difference

TSS 1479 265 -1214

TP 33 21 -12

BOD** 2691 145 -2545

TKN 930 192 -738

NH3 609 93 -516

NO2- NO3 51 516 465

*Based on average daily concentration Aug - Dec 2010

**2006 BOD is TBOD, 2010 BOD is CBOD

Conclusions

• The BAF process itself is complex, while at the same

time, simple to operate. Plant operators do not have

to concern themselves with conventional activated

sludge process parameters such as sludge return

and wasting rates, sludge bulking, sludge washout,

etc.

• Attention to backwash is critical to performance of

BAF operation

• Changing BAF cell run time from fixed time to solids

loading based time resulting in significant

improvements in LRWRP effluent quality.

Questions

Paul Drca

Manager Environmental Quality

City of Windsor

519-253-7217