Content uploaded by Benoit Barbeau
Author content
All content in this area was uploaded by Benoit Barbeau
Content may be subject to copyright.
1
1
Performance of an hybrid
membrane process using
biological PAC
Benoit Barbeau, Simon Léveillé, Sébastien
Charest and Annie Carrière
École Polytechnique de Montréal
Alain Gadbois
John-Meunier Inc. – Veolia
Membrane Technology conference, March 28, 2011
2
Background
At the end of the 1990s, nanofiltration (NF) was
seen as the most promising technology for
drinking water treatment.
This process has not been as largely adopted due :
Energy costs
Fouling issues (requires a pretreatment)
Hybrid membrane processes integrating both PAC
adsorption and low pressure membranes (UF or
MF) offers an alternative to NF.
3
Integrating activated carbon with
membranes: the North American experience
Ultra(Micro)filtration + O3+ GAC filters examples:
Twin Oaks, CA (2007): 400 MLD
•UF + O3+ GAC
Lakeview,Ont. (2008): 363 MLD
•O3+ BAC + UF
Filters are expensive to build
PAC is more efficient than GAC
Integrating PAC with UF is therefore
of great interest
Twin Oaks
Lakeview
Source: Ge-Zenon
Source: Ge-Zenon
4
Immersed membrane configuration
The challenge of fouling
What are the impacts of operating
conditions on fouling and process
performance ?
Can the process be operated in
biological mode, i.e. leaving the PAC
in the reactor?
5
Project description
PHASE 1: Mini-Pilot PAC bioreactors (1 L) fed
during 160 days with filtered-ozonated
surface waters (2007-2009)
Impact of PAC concentration
Impact of PAC diameter
PHASE 2: Industrial pilot (50 m3/d) (2009-ongoing)
Settling + Ozonation + PAC/immersed UF
Impact of flux on fouling
Performance for NH3, THM, BDOC
6
Phase 1: Mini-pilot study
Description of the bioreactors
Effluent
Influent (ozonated filtered water)
Purge
Effluent
Stirred 1-L
bioreactor
10 µm sieve
7
Under stable biological mode (100-160 d of operation)
Increasing PAC concentration from 5 to 25 g/L significantly
increased performance.
Average performance: Reduction of 9%, 43% & 93% of DOC,
BDOC and NH3 using 25g/L of PAC
Equal perfomance using a PAC of 25 µm or 200 µm.
Increasing contact time over 30 min does not significantly
improve performance.
Biological PAC tends to form agregates
Phase 1: Mini-pilot study
Main conclusions
8
1: Settling (Actiflo®) (8 feet)
1 2 34
2: Ozonation: 1.5 mg O3/L (8 feet)
3: Pressurised HMP (Opaline S®) (12 ft)
4: Immersed HMP (Opaline B®)(10 ft)
Phase 2: Industrial Pilot
Description
9
Puronmembrane
Hollow fiber (PES)
Porosity: 0.05µm(ultrafiltration)
Membranearea: 10m²
Configuration: Immersed
PICAHYDROLP39
(Woodbase)
Size: D50 :18.7microns
Type:
Membranesspecifications
Technology:
PACspecifications
Flux: 15to25LMH
Concentration: 10‐12g/L
Phase 2: Industrial Pilot
Bioreactor specifications
300LMH
Aeration rate:
5minfiltration+
1minbackwash
Filtrationcycle:
Backwash rate: 30LMH
Pretreatment: Sludge blanket clarifier
10
Research Objectives
Evaluate the viability of the Opaline-B
operated under biological mode
-Two parallel reactors
With or without PAC
-Side comparison with the existing plant
Pilot located at the Ste-Rose WTP
Settling + Filtration + Ozone + BAC filters
Source water: High TOC (6-8), ammonia (200-
300 ppb), E. coli = 100-500 CFU/100mL
11
Pilot Plant Study
Experimental phases
Date PAC age [PAC] Flux
(L/m2/h) Comments
Phase
I
13july‐
19oct. 2009 0‐ 120d20g/L 45 à15 Excessivefouling
Phase
II
18Feb.‐
4May2010 200‐ 280 d 12g/L 15à25 New membranes
Phase
III
13May‐
ongoing
Fixed:
‐30 d
‐60d
10g/L 25
Same membranes
New PAC
12
0
10
20
30
40
50
0 0.5 1.0 1.5 2.0 2.5
Flux(L/m2/h)
Transmembrane presssure (psi)
Selection of the Operating Flux
Technique used: « flux stepping »
0
100
200
300
08:30 09:30
Flux stepping
heure
Flux (L/h)
36
25
The presence of biological PAC reduces the critical flux.
Opaline w/o PAC Membrane w/o PAC
Opaline with PAC
13
Critical Flux
W/O PAC: 36 LMH at 7 oC (48 LMH at 20 oC)
With PAC: 25 LMH at 7 oC (37 LMH at 20 oC)
Selected Flux:
Initially 15 LMH; increase to 25 LMH
5 min filtration + 1 min backpulse at 30 LMH
Available Hydraulic Residence Time
92 min at 15 LMH
55 min at 25 LMH
Selection of the Operating Flux (cntd)
14
Results:
Permeability
0
100
200
300
400
500
600
700
800
900
1000
Operating days
with PAC
w/o PAC
15LMH
CIP CIP
Permeability @20oC
(LMH/bar)
15
Results:
Suspended Solids inside the Bioreactor
0
2
4
6
8
10
12
14
16
18
20
MES(g/L)
OpalineavecCAP
MembranesansCAP
Phase II Phase III
Physical
cleaning
PAC
dosing
Physical
cleaning
+3 g/L
with PAC
without PAC
16
Physical fouling
MinimalPAC
accumulation,mainly
at thelower endof
themodule
Theairinjectionat the
bottom minimizes PAC
settling
17
Chemical cleaning efficiency
Flux test for new membrane and after chemical cleanings
-3,50
-3,00
-2,50
-2,00
-1,50
-1,00
-0,50
0,00 0 20406080
TMP (psi)
Normalized flux @20C (LMH)
CW-new
CAP 200d-new
CW-3months
CAP 0d-3months
CW-6months
-3,50
-3,00
-2,50
-2,00
-1,50
-1,00
-0,50
0,00 0 20406080
TMP (psi)
Normalized flux @20C (LMH)
CW-new
CW-3months
CW-6months
Membrane with CAP Membrane W/O CAP
18
Process performance
Ammonia
NH4+removal (%)
0
20
40
60
80
100
Phase II Phase III
[NH4+] affluent:
50 to 300 µg/L
1 to 15 oC15 to 30 oC
Opaline(with CAP)
Membrane(w/oCAP)
BACeffluent
19
Hypothesis for nitrification decline
Increase metals content of PAC
Old PAC - 240 DAYS
PAC - Virgin
Aluminium
Calcium
Fer
Carbone
Autre(O,N,P,S)
Iron
Others
Alum flocs accumulation may inhibit nitrification
20
PAC texture changed over time. This may hinder O2and/or
NH4+diffusion to AOB.
Biological
PAC New
PAC
Hypothesis for nitrification decline
PAC floculation/agregation
21
Process performance
Dissolved organic carbon (DOC)
1,0
1,5
2,0
2,5
3,0
3,5
DOC(mg/L)
Phase II Phase III
1 to 15 oC15 to 30 oC
Affluentpilot
AffluentBAC
No pilot O3
Membrane(w/oCAP)
Opaline(with CAP)
BACeffluent
22
Process performance
BDOC
Phase II Phase III No pilot O3
0,00
0,10
0,20
0,30
0,40
0,50
0,60
0,70
0,80
0,90
1,00
BDOC(mg/L)
AffluentBAC
Affluentpilot
Opaline(with CAP)
Membrane(w/oCAP)
BACeffluent
23
Uniform formation conditions (UFC)
Cl2res. = 1ppm @ 24h; pH=8; T=22oC
Process performance
THMs
Sampling campaign
THM–UFC(µg/L)
8 may 10
20 feb.10
20 jul. 09
0
10
20
30
40
50
60
70
80
90
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
No pilot O3
AffluentBAC
Affluentpilot
Opaline(with CAP)
BACeffluent
Membrane(w/oCAP)
Phase II Phase III
Phase I
24
0
5
10
15
20
25
30
THM/DOC (µg/mg)
DOC
2.74 DOC
2.62
DOC
2.24 DOC
2.63 DOC
2.25
Process performance
Specific THM formation
Despite lower
DOC levels,
the reactivity
of the
accumulated
OM is greater
(21 vs 15 µg/mg)
W/O PAC
With PAC
Influent Effluent
Influent
Full-scale
BAC
Membrane
bioreactors
25
Conclusions
Combining biological PAC with UF
Lowered the critical flux by 30%
Caused reversible fouling which requires attention
(cleaning and thoughtful aeration…) and irreversible
fouling
However, filtration cycles >3 months at 25 LMH
Allowed complete nitrification at 6oC as long as the
PAC is not too old (or too much contaminated by alum
flocs carry over)
Provided equivalent or better water quality than the
plant (DOC, BDOC, THM) (both 30d and 60d)
Allowed for more operation flexibility (adsorption vs
biodegradation)
26
Conseil de recherches en sciences
naturelles et en génie du Canada
www.polymtl.ca/chaireeau