3/23/10
1
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 1 CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 2
Outline
! Introduction
• PAT in Raw-Materials – i.e., Oils Qualification
• PAT in Production – i.e., Process Supervision
• PAT in Final Product – i.e.,Biodiesel QC
! Conclusions
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 3
PAT Building Blocks
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 4
PAT Based Biofuel Production
! Assessing Raw-Materials' Variability – Complex raw-materials (RM): fast qualification schemes (QC) – Opportunity for RMs supervisory systems (QA) – Formulating mixed lots of RMs for targeted specifications
! Handling RM Variability During Processing – Monitoring quality attributes in-process (IPCs) at-line or on-line – Process supervision (e.g., multivariate trajectories) – Controlling critical to quality parameters
! End Product Quality Control – Biofuels multiparametric QC (e.g., conformity index) – Blending operations for targeted specifications
3/23/10
2
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 5
raw-mat process product
investment costs (storage) & profit margins (QC)
a batch process likely to become continuous prone to process spectroscopy monitoring
catalytic reaction phase separation
washing & drying biodiesel
Vegetable Oils (soybean, palm, jatropha etc)
Algae Oils Waste frying Oils
Animal Fats
Biodiesel
How can PAT be important ?
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 6
Monitoring with a Lab Instrument
Interfaced with the Process
Process Finger-printing Process Trajectories Process Supervision, Fault Detection End-Point Determination Conformity Index (Multiparametric QC)
Accurate Calibration Models between spectral and analytical data - methyl ester content - residual alcohol content - water ...
NIR Transflectance Probes (Solvias, AG)
ABB MB160 FT-NIR
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 7 CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 8
Outline
! Introduction
• PAT in Raw-Materials – i.e., Oils Qualification
• PAT in Production – i.e., Process Supervision
• PAT in Final Product – i.e.,Biodiesel QC
! Conclusions
3/23/10
3
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 9
Raw Materials Qualification
Type of Oil PCA of Pure oils
Palm
Rapeseed
Soybean
Iodine number
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 10
Raw Materials Monitoring Type of Oil/Blendings PCA of Mixtures of oils
Palm
Rapeseed
Mixtures
Iodine number
Soybean
EN14214 imposes Iodine Number < 120
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 11
Raw Materials Monitoring QC of Oil Properties NIR PLS Calibrations
Property Error of the reference method
Error of NIR models
(RMSEP)
Calibration Range
Water content (mg/kg)
50 68 478 – 1776
Iodine Value (g I2/100 g)
3 1.0 60 – 126
Acid Value (mg KOH/g)
0.1 0.41 0.13 – 6.56
Dynamic Viscosity (cP)
2 1.9 5 – 74
Density at 15ºC (kg/m3)
0.5 0.9 875 – 925
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 12
Raw Materials Monitoring QC of Oil Blending (%) NIR PLS Calibrations
Property Error of NIR
models (RMSEP in %)
Calibration Range
% Soybean Oil (SO) 2.4 0 -100%
% Rapeseed Oil (RO) 3.2 0 -100%
% Palm Oil (PO) 1.2 0 -100%
Note: oils mixtures were laboratory prepared by weighing and mixing the oils in the desired proportions
3/23/10
4
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 13
Raw Materials Formulation Oil Mixtures Compositions (% of SO in SO/RO/PO mixtures)
NIR PLS Calibrations
(× = calibration set; ▲ = validation set). CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 14
Outline
! Introduction
• PAT in Raw-Materials – i.e., Oils Qualification
• PAT in Production – i.e., Process Supervision
• PAT in Final Product – i.e.,Biodiesel QC
! Conclusions
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 15
Following the transesterification reaction
Production Process Monitoring
Raw-‐materials
In process samples
Biodiesel
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 16
Production Process Monitoring
Following Biodiesel/Glycerol Separation
Batch Process
NIR Probe in Biodiesel Phase
Reaction End (stirring On)
Phase Separation (stirring Off)
Phase Equilibrium (stirring Off)
3/23/10
5
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 17
Outline
! Introduction
• PAT in Raw-Materials – i.e., Oils Qualification
• PAT in Production – i.e., Process Supervision
• PAT in Final Product – i.e.,Biodiesel QC
! Conclusions
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 18
NIR Fingerprinting- Different Biodiesel Lots
Final Product Quality Control
IV 110 – 120
(opt
imum
)
IV 60 -‐ 6
5
IV 65 -‐ 110
IV 12
0 -‐ 135
Biodiesel before refining
Refined Biodiesel
IV-‐ Iodine Value
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 19
Accurate calibrations developed for 15 quality biodiesel quality specs
(EN 14214 25 parameters)
Property EN 14214 limit
Calibration Range
Error of NIR models
(RMSEP)
Error of the reference method
Methyl esters 1.5 0.9 99.3 – 78.4 % 96.5 < )%(
Monoglycerides (mg/kg)600 654 11738 – 2641 8000 >
Diglycerides (mg/kg) 184 389 7846 – 82 2000 >
Triglycerides (mg/kg) 282 - - 2000 >
Free Glycerol (mg/kg) 56 - - 500 >
Total Glycerol (mg/kg)218 178 4230 – 1371 2500 >
Methanol (mg/kg) 70 67 2864 – 2 2000 >
Linolenic acid methyl ester (C18:3) (%) 1 0.18 8.9 – 0.3 12 >
Final Product Quality Control
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 20
Property EN 14214 Limit
Calibration Range
Error of NIR models
(RMSEP)
Error of the reference method
Methyl esters composition (%) Not specified Depends on
the ester Depends on
the ester Depends on
the ester
Iodine value (g I2/100 g) 3 0.7 135 – 62 120 >
Density at 15ºC (kg/m3) 0.5 0.9 925 – 875 900 – 850
Kinematic viscosity at 40ºC (mm2/s) 0.08 0.08 4.91 – 3.67 5.0 – 3.5
Water content (mg/kg) 50 72 1859 – 218 500 >
Acid value (mg KOH/g) 0.05 0,05 0.68 – 0.09 0.5 >
Oxidative Stability (h)
6 4.32 – 16.8 0.28 0.12
CFPP (ºC) Varies -13 to 5 1.0 1
CP (ºC) Not specified -3 to 9 0.4 1
Final Product Quality Control
3/23/10
6
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 21
EN 14214 vs ASTM
Critical Specification Quality Parameters According to BQ 9000
Final Product Quality Control
Property EN 14214 limit
ASTM 6751 limit
NIR Calibration
Methanol Content √ 0.2 > 0.2 > )%(
Water & Sediment (% Vol.) - 0.05 > -
Water Content √ - 0.05 > )%(
Sulfur (mg/kg) or 500 X 15 10 >
Free Glycerol X 0.02 > 0.02 > )%(
Total Glycerol √ 0.24 > 0.25 > )%(
Acid number (mg KOH/g)
< 0.5 < 0.5 √
Oxidation Stability (h)
> 6 > 3 √ Cloud Point
(º C) Contract Contract √ CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 22
Biodiesel Quality Control
Biodiesel analyses are very expensive and time consuming
It takes + 2 days to perform by the reference methods the analyses of the 15 quality parameters calibrated with NIRS
To perform the simultaneous analyses of the calibrated properties by NIRS, it is only necessary to acquire the NIR spectrum of the sample, a procedure that takes less than 2 minutes
The prediction errors of NIRS models are comparable to the reference methods errors
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 23
Good Chemometric Practices Robust long-term accurate calibrations established using industrial and lab-scale samples
Methanol
(x calibration set; ▲ validation set)
Example: methanol and FAME in biodiesel
FAME
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 24
Mixtures
of oil,
biodiesel
and
fossil diesel Diesel
Biodiesel
Oil
End-Product QA: Blending & Counterfeit
3/23/10
7
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 25
Outline
! Introduction
• PAT in Raw-Materials – i.e., Oils Qualification
• PAT in Production – i.e., Process Supervision
• PAT in Final Product – i.e.,Biodiesel QC
! Conclusions
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 26
! Assessing Raw-Materials' Variability – Qualification (QC) of Complex Raw-Mat – Opportunity for Raw-Mat Supervisory Systems – Formulating Oil-Mixes for targeted specifications
! Handling RM Variability During Processing – Monitoring IPCs at-line or on-line – Process supervision (e.g., mapping process trajectories) – Controlling critical to quality parameters
! End Product Quality Control – Biodiesel multiparametric QC (e.g., conformity index) – Blending operations for targeted specifications
Conclusions
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 27
Conclusions
NIR Spectroscopy is a very powerful analytical technique in biodiesel production specially if used as
process analytical technology
fingerprinting, multiparametric and whole-process-based
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 28
Conclusions
NIR Spectroscopy is a very powerful analytical technique in biodiesel production specially if used as
process analytical technology
fingerprinting, multiparametric and whole-process-based
NIRS is accepted as analytical technique for parametric release of
food products (that we eat – when healthy) pharmaceuticals (that we take – when ill) why not for Biodiesel ? (oh no, not in my car !)
3/23/10
8
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 29
Breewing & Dairy Biomanufacturing
Pharma Petrochemical
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 30
4TUNE Engineering Ltd better processes, better products
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 31
Offices in Lisbon, Clients Everywhere a charming history-rich European capital
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 32
Biotech
Chemical (CPI)
Pharma (API)
Petrochemical
Biofuels
Food
3/23/10
9
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 33 CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 34
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 35 CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 36
Thank you !
3/23/10
10
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 37
Recent Papers by Us
P. Felizardo, P. Baptista, J.C. Menezes and M.J. Neiva Correia, Predicting methanol and water content in biodiesel by near infrared spectroscopy, NIR news 19(3), 7–9, 2008.
Baptista P., Felizardo P., Menezes J.C., Correia M. J. N., Multivariate Near Infrared Spectroscopy Models for Predicting the Methyl Esters Content in Biodiesel, Analytica Chimica Acta 607, 153-159, 2008.
Felizardo P., Baptista P., Menezes J.C., Correia M. J. N., Multivariate Near Infrared Spectroscopy Models for Predicting Water and Methanol Content in Biodiesel, Analytica Chimica Acta V. 595, Issues 1-2, 107-113, 2007.
Baptista P., Felizardo P., Menezes J.C., Correia M. J. Neiva, Monitoring the Quality of Oils for Biodiesel Production Using Multivariate NIR Spectroscopy Models, J. of Near Infrared Spectroscopy, 16, 445-454, 2008
Baptista P., Felizardo P., Menezes J.C., Correia M. J. Neiva, Multivariate Near Infrared Spectroscopy Models for Predicting the iodine value, CFPP, Kinematic Viscosity at 40ºC and Density at 15ºC of Biodiesel, Talanta, 77, 144-151, 2008
J.C. Menezes, P. Felizardo, M.J. Neiva Correia, The Use of Process Analytical Technology (PAT) Tolls in Biofuels Production, Spectroscopy News, Sept. , 2008
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 38
CPAC SATELLITE WORKSHOP, March 22-24, 2010 – Rome (Italy) 39