WA2 Hybrid ceramic membrane filtration in water treatment Erwin WA2.pdf · WA2 Hybrid ceramic...
Transcript of WA2 Hybrid ceramic membrane filtration in water treatment Erwin WA2.pdf · WA2 Hybrid ceramic...
WA2 Hybrid ceramic membrane filtration in water treatment
This project has received funding from the European Union’s Seventh Programme for Research, Technological Development and Demonstration under Grant Agreement no. 308339.
Final meeting, June 17+18 2015 Erwin Beerendonk, Egbert Dubbelink, Thomas Wintgens, Arslan Ahmad, Andreas Nahtstedt, Jörg Gebhardt, Patrice
Nagtegaele and Marie-Pierre Denieul
• Introduction • Work packages • Conclusions • Success stories, leassons learned and
challenges
23.06.2015 2
23.06.2015 5
Introduction
• UF (poresize >5 nm - 0,1 μm) and MF (>0,1 μm) will not/partly remove emerging compounds and pathogens
• Hybrid systems – Coagulation to improve removal of NOM and viruses (MF)
and improve operation – Ion exchange to improve removal of NOM and operation – Activated carbon (PAC) or ozone to improve removal of
emerging pollutants
CMF = High investment costs!
WP21 Ceramac (RWB)
WP22 Hybrid Ceramic Membrane Systems (KWR)
WP23 Neural Net Control (IWW) WP24 Laser-Induced Breakdown Detection (Cordouan)
WP21 Ceramac (RWB)
WP22 Hybrid Ceramic Membrane Systems (KWR)
WP23 Neural Net Control (IWW) WP24 Laser-Induced Breakdown Detection (Cordouan)
•Innovation in reactor design (200 elements in 1 vessel, reduce footprint, piping etc)
•Further optimizing design
•Pilot testing and implementation full scale PWN
WP21 Ceramac (RWB)
WP22 Hybrid Ceramic Membrane Systems (KWR)
WP23 Neural Net Control (IWW) WP24 Laser-Induced Breakdown Detection (Cordouan)
•High potential (Techneau) •Optimizing performance HCMS
•Compare with state of the art
•Demonstration in Hamburg (HSE)
WP21 Ceramac (RWB)
WP22 Hybrid Ceramic Membrane Systems (KWR)
WP23 Neural Net Control (IWW) WP24 Laser-Induced Breakdown Detection (Cordouan)
•Developed in EU-Life “Purifast”
•Adapt to full scale
•Training neural net
•Pilot and full scale UF backwash water WAG
WP21 Ceramac (RWB)
WP22 Hybrid Ceramic Membrane Systems (KWR)
WP23 Neural Net Control (IWW) WP24 Laser-Induced Breakdown Detection (Cordouan)
•Demonstrate nanoparticles analyzer on lab, pilot and full scale
•LIBD prototype mobile system
•Field demonstration LIBD
23.06.2015 11
WP21 Optimisation Ceramac® concept (1)
• Design 192 membranes in one reactor • Tests, validation and development of the top
and bottom plate and hydraulic behavior • Validation tests May 2013 until September
2013 on the full scale pilot • The top and bottom carbon fiber plate
concept is tested in a separate test unit
23.06.2015 12
WP21 Optimisation Ceramac® concept (2)
• Test of 793 backwashes (critical) showed stable process and applicable to the full scale plant
• Successful implementation of the full scale CeraMac® in Andijk 3 (Capacity and quality meet design criteria)!
• Application of HCMS improved (footprint, costs, sustainability and full scale reference)
Pilot locations • WWTP Almelo
(The Netherlands) • WWTP Basel
(Switzerland)
23.06.2015 13
WP22 HCMS to increase application CMF (1)
23.06.2015 14
PAC-CMF pilot at WWTP Almelo
WP22 HCMS to increase application CMF (2)
• Experiments: (1) OMP removal (2) Operational stability • WWTP effluent + OMPs = Feed pilot plant. • Cocktail of OMPs dosed, each ≈1µg/L. • PAC dose (mg/L): 0, 15, 30, 60 [precoat mode] • BW: pressurized (5 bar) with permeate and air • CEB: BW with permeate and NaOCl
23.06.2015 15
OMPs dosed µg/L metoprolol 0,936 gemfibrozil 0,953
sotalol 0,901 tramadol 0,951
carbamazepine 0,962 venlafaxine 0,881 diclofenac 1,027 atenolol 0,972
propranolol 0,849 trimethoprim 0,973
sulfamethoxazool 0,980 ketoprofen 0,999 bezafibraat 0,918
diatrizoic zuur 1,019 metronidazole 1,005
fenazon 0,996 cyclophosphamide 0,797
pentoxifylline 0,970
Filtration time 15 min Filtration flux 60,80,100, 120 L/(m2·h) BW frequency 4 times per hour BW time <5 sec CEB frequency 1 time per 6 hour Chemicals used NaOCl (12,5 wt%) Soaking time 5 min
WP22 HCMS to increase application CMF (2)
23.06.2015 16
0 0,2 0,4 0,6 0,8
1 1,2 1,4 1,6 1,8
20 30 40 50 60 70 80 90 100
TMP
(bar
)
Time (hour)
TMP with 30mg/L PAC Flux=100 L/m2.h
TMP
WP22 HCMS to increase application CMF (2)
Removal of OMPs
23.06.2015 17
WP22 HCMS to increase application CMF (2)
Pilot tests in Basel (double HCMP) • Combination of PAC sorption and UF-filtration in one process:
HCMP (Hybrid ceramic membrane process)
• Two parallel systems
Double HCMP system
Membrane ITN Nanovation
Membrane area 0.8 m2
Contact zone 35 L
Filtration tank 150 L
PAC dosage 20 mg/L
Fe dosage 4 mg/L
Feed per line Ca 20 L/h
WP22 HCMS to increase application CMF (2)
Experimental set-up (double HCMP)
• HCMP (Hybrid ceramic membrane process)
• Two parallel systems
Double HCMP at pilot hall Top view on flat sheet ceramic membranes
WP22 HCMS to increase application CMF (2)
Experimental setup
Experiment
• HCMP operation on WWTP effluent as reference (without PAC)
• PAC addition
• PAC/Fe3+ addition
• Optimization for stable operation (filtration flux)
• Investigate chemical cleaning strategy
WP22 HCMS to increase application CMF (2)
Membrane performance
• Filtration of WWTP effluent
• Comparability of the 2 HCMPs (same operational parameters)
Operational parameters Flux: 25 LMH CEB frequency 2.5 d PAC: 0 Fe3+: 0
WP22 HCMS to increase application CMF (2)
Operational parameters Flux: 25 LMH CEB frequency 2.5 d PAC: 20 mg/L Fe3+: 0
Micropollutants- concentration and removal with PAC addition
• Good reduction of non-polar substances
n=11
WP22 HCMS to increase application CMF (2)
Membrane performance with PAC + Fe3+ addition
• PAC dose: 20 mg/L in both HCMP
• Fe dose: 4 mg/L - in one HCMP only
• 1st experiment: little difference between the two HCMPs until problems with aeration in HCMP 1
Operational parameters Flux: 25 LMH CEB frequency 2.5 d PAC: 20 mg/L Fe3+: 4 mg/L (only HCMP 1)
(with Fe3+) (no Fe3+)
wrong flux setting
uneven aeration in HCMP1
WP22 HCMS to increase application CMF (2)
Membrane performance with PAC + Fe3+ addition
• PAC dose: 20 mg/L in both HCMP
• Fe dose: 4 mg/L - in one HCMP only
• 2nd experiment: little difference between the two HCMPs, until the dosing point of Fe3+ moved
from contact zone into the membrane tank
Operational parameters Flux: 18 LMH CEB frequency 1 d PAC: 20 mg/L Fe3+: 4 mg/L (only HCMP 2)
(with Fe3+) (no Fe3+)
Point of Fe3+ dosing changed
WP22 HCMS to increase application CMF (2)
• Removal increases with PAC dose, as expected
• At 30mg/L PAC dose and flux=100 lmh, HCMF process remained stable
• PAC-CMF applicable on waste water (calculation of costs and sustainability to be checked (WA5))
23.06.2015 25
Conclusions
WP22 HCMS to increase application CMF (2)
23.06.2015 WP2.3 26
WAG Wassergewinnungs- und -aufbereitungsgesellschaft Nordeifel (treatment plant Roetgen)
Pilot UF treating backwash water process (Sept. 2014 – Feb.
2015) Two treatment lines :
- UF-modules Lab 1.5 MB 1.0 (Multibore®)
- Membrane area 1m² (small area / full length)
- Capillaries per fiber 7 - Inner diameter 1.5 mm - Outer diameter 6.0 mm - Pore size ca. 0.02 µm - Material: PESM
WP23 Process control with ANCS (2)
27
Pre-Filtration
Kalkstein-Filtration
Disinfection
Storage
Flocculation (in-line)
Ultrafiltration 1st stage
PAK (if necessary)
NaOH / CO2
Al2(SO4)3
CO2
NaOH (if necessary)
Backwash Water
Recycling Ultrafiltration
2nd stage without
chemicals
River
Neutralization
with chemicals
Raw water: Dreilägerbach Reservoir
with chemicals
WP23 Process control with ANCS (3)
23.06.2015 WP2.3 28
ANCS-Structure: − Clustering of input data with similar patterns
(feed water quality, parameter settings) − Weighting input channels for single ANCS
neurons to link data clusters with their effect on treatment efficiency
WP23 Process control with ANCS (4)
UF100 UF200
Gemessen Optimiert Einsparung/ Verbesserung
Gemessen Optimiert Einsparung/ Verbesserung
Elektrische Energie Kosten [€/m³]
1,901 1,671 0,23 2,521 2,053 0,467
Erlös [€/m³] 0,642 0,649 0,006 0,653 0,659 0,006
Energieeinsparung von 12,10% und eine Verbesserung der Produktivität um 0,98%
Energie Einsparung von 18,55% und Verbesserung der produktivität um 0,97% .
Optimisation results
Ergebnisse im Überblick
• Modelling of permeabilility after/before backwash • Improving permeability results in reduction of energy
use • May result in decrease of production optimise to
find the best compromise
WP23 Process control with ANCS (5)
Conclusions • Succesfull modelling of start- and end-permeabibility
after/before backwash • ANCS results in cost reduction of 4 - 25 %
• ANCS proven on membrane application (full scale to be
done), before already on coagulation/flocculation and distribution
WP23 Process control with ANCS (6)
Context & Objectives • Context:
• Current tests for checking UF membrane integrity (PDT / DAF) involve: • Reuse of produced water • Interruption of operation of membrane units • Loss in production + Costs $
• Objectives: Evaluation of potentials of the LIBD technology at lab- & pilot scale • Online semi-quantitative detection of nano particles (NP) in water to
ensure early monitoring of loss of integrity of UF membranes (skid or module)
23.06.2015 WA2.4 progress - Process monitoring with LIBD 32
Need for online monitoring of the integrity of UF membranes designed for the Drinking Water treatment market
WP24 Process monitoring with LIBD (1)
23.06.2015 WA2.4 progress - Process monitoring with LIBD 34
The challenge
A perfect method for low-pressure membrane integrity test should be:
on-line, continuous, easy to handle, reliable, highly
sensitive & low cost
Magellan Technology - LIBD
23.06.2015 WA2.4 progress - Process monitoring with LIBD 37
+ Robustness
23.06.2015 WA2.4 progress - Process monitoring with LIBD 38
The MAGELLAN analyzer is compact, transportable and robust
Test performed in Magellan Vial trials
• SYNTHETIC WATERS spiked in Flow though cell (MQ UPW) • Validate the field of measuring range , check the repeatability of the measurement, validate the detection of different
natures particles ORGANIC: PS 20 nm / 40 nm / 100 nm (concentration range : 103 – 107 part/mL) INORGANIC: AG 20 nm / 100 nm (concentration range : 103 – 107 part/mL) BIOLOGIC: MS2 phages 27 nm (concentration range : 104 – 108 part/mL) * SCREENING EFFECT : Mix PS 20nm + 40nm * NATURE EFFECT Mix PS 20nm + AG 20nm * TEMPERATURE EFFECT
• NATURAL WATERS in Flow though cell => Verify the feasibility of the measurement of a real water
Membrane pilot
23.06.2015 WA2.4 progress - Process monitoring with LIBD 41
Description Concentration of nanoparticles spiked Aim
Trial 1 Reference blank test with ultrapure water (UPW)
No Reference blank for each type of membrane
Trial2 MS2 phages in UPW 108 PFU/ml MS2 phages
To compare the results done by LIBD technology with the Plaque Forming Unit counting method.
Trial 3 Synthetic water spiked 6. 108 part/ml PS 20 nm and PS 40 nm 6.107 part/ml PS 100 nm and 6.108 PFU/ml MS2
phages
To test different size of nanoparticles. Evaluate the membrane cut-off.
Trial 4 Natural water No Study of effect of the water matrix
Trial 5 Natural water spiked PS 20 nm 1010 part/ml PS 20 nm
+ Repeatability
23.06.2015 WA2.4 progress - Process monitoring with LIBD 42
ORGANIC (PS)
UPW INORGANIC (Ag)
BIOLOGIC (MS2 phages)
Good repeatability regardless
the nanoparticle type
+ Reproducibility
23.06.2015 WA2.4 progress - Process monitoring with LIBD 43
UPW ORGANIC (PS)
Good reproductibility regardless the nanoparticle type
+ Sensitivity (concentration range)
23.06.2015 WA2.4 progress - Process monitoring with LIBD 44
• Organic (PS 20 nm) • Organic (PS 40 nm)
Order of magnitude for: • Detection: 103 part./mL • Quantification: > 106 - 107 part./mL
ZOOM IN
ZOOM IN
+ Detection of different types of particles • Biologic (MS2 phages 27 nm) • Inorganic (Ag 20 nm)
23.06.2015 WA2.4 progress - Process monitoring with LIBD 45
The LIBD is able to detect different types of nanoparticles with more or less sensitivity (no differentiation according the nature of particles)
+ Capacity to detect a loss of membrane integrity: evaluation of screening effect
23.06.2015 WA2.4 progress - Process monitoring with LIBD 46
Conclusions
23.06.2015 WA2.4 progress - Process monitoring with LIBD 52
Advantages
Repeatability
Reproductibility
No temperature effect
Integrity monitoring
Cut-off evaluation
Detection mode
Limits of technolgy
Quantification concentration
limit
Screening effect
Flow though cell fouling
Software
No Log removal quantification
LIBD : Demonstrates good repeatability and reproductibility Allows to detect different nature of particles but without differentiation Detects nanoparticles down to 103 part./mL but able to quantify
concentrations from 106 - 107 part./mL only Is not influenced by temperature effect and transport Can detect losses of integrity as well as damage levels Cannot monitor the log removal (screening effet & concentration level) Can be used to detect nanometric breaches in permeates in detection
mode No operational analysis software for small sizes up to now
Online trials should allow to check: Possible contamination of the flow through cell Influence of bubbles or flow rate variations Robustness in the harsh environment of water works (vibrations, noise,
temperature, hymidity,…) By-pass or pipeline implementation during the cleaning or backwashs
NEW TECHNOLOGY NEW WAY OF WORKING
• WA4: Evaluating toxicity before and after PAC-CMF (testing Almelo)
• WA5: Evaluating drivers and barriers of HCMS and ANCS (workshop Roetgen, June 11)
23.06.2015 53
Synergy with other WA’s
23.06.2015 54
Success stories (1)
– Ceramac • Improved footprint, costs and sustainability • Full scale reference PWN • Demo plant PUB Singapore • Pilot + demo plant SWW (UK) • Pilot Cairns (Australia) • Pilot Manitowoc (USA)
23.06.2015 55
Success stories (2)
– HCMS: PAC-CMF increased performance and removal of emerging compounds in waste water treatment
– ANCS • Reduce costs ultrafiltration (full scale UF Roetgen to be
finished) • Other drinking water plants and treatment
processes(coagulation, distribution)
Thank you for your attention!
This project has received funding from the European Union’s Seventh Programme for Research, Technological Development and Demonstration under Grant Agreement no. 308339.