[PPT] - Courtney Phillips Leap Technologies John Edwards Process NMR PowerPoint Presentation

SLIDE 1

High Throughput Petroleum Stream Analysis in Refinery Process Laboratories: Benchtop NMR Offers Timely Results with Automation & Chemometrics

Courtney Phillips Leap Technologies John Edwards Process NMR Associates

SLIDE 2

Residual Fluidized Catalytic Cracking Feed-stream Analysis Traditional Analysis – Refractive Index, Distillation,Viscosity Specific Gravity Calculation – Watson K-Factor Outcome: aromatic carbon number aromatic hydrogen number total hydrogen content NMR Proposition: Detailed hydrocarbon analysis for kinetic model development

SLIDE 3

SLIDE 4

CH3 CH2 Monoaromatic DiAromatic Di+Tri Aromatic Tri+Tetra Aromatic Alpha Aromatic CH2 CH3 Ar-CH2-Ar

SLIDE 5

Data Stacked – Referenced – TMS excluded – spectrum binned (integrated every 0.04 pm from -1 to +11 ppm) - Normalized

0.9399 0.9574 0.9230 0.9255 0.9354 0.9122 0.9000 0.8987 0.8823 0.9786 0.9112 0.9 0.8781 0.9270 0.9255 0.9359 0.9006 0.8745 0.9271 0.9215 0.8882 0.9233 0.8802 0.8847 0.8842 0.8628 0.8753 0.9217 0.9208 0.9224 0.8888 0.9258 0.8795 0.8663 0.8990 0.8757 0.9488 0.9272 0.9021 0.8762 0.9233 0.9390 0.8916 0.9042 0.9213 0.8829 0.8779 0.8930 0.9130 0.9208 0.9218 0.8979 0.9364 0.8946

X-Block Y-Block PLS

SLIDE 6

50 100 150 200 20 40 60 80 100 120 140 160 Variables Data

60 80 100 120 140 160 180 200 220 1 2 3 4 5 6 Variables Data

Data after importation into Eigenvector Research Chemometrics Package – Note Reversal of Spectrum

SLIDE 7

10 20 30 0.5 1 1.5 2 2.5 3 Hotelling T^2 (99.94%) Q Residuals (0.06%) 0.2 0.4 0.6 0.8

3
2
1

1 2 3 Leverage Y Stdnt Residual 1 0.85 0.9 0.95 1 0.85 0.9 0.95 1 Y Measured 1 Y CV Predicted 1

100
50

50

20
10

10 20 Scores on LV 1 (89.72%) Scores on LV 2 (3.37%)

Samples/Scores - PLS 9 LVs - Binned Data.csv, Density.xlsx

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 07-Oct-2014 23:15:09.003 Author: John@JCEPNA X-block: Binned Data.csv 46 by 248 Included: [ 1-6 8-11 14-32 36-38 41-46 48-54 56 ] [ 1-248 ] Preprocessing: Mean Center Y-block: Density.xlsx 46 by 1 Included: [ 1-6 8-11 14-32 36-38 41-46 48-54 56 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 9

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.0034 g/ml RMSECV: 0.0052 g/ml Bias: 3.33067e-016 CV Bias: -9.35779e-005 R^2 Cal: 0.980 R^2 CV: 0.954 SSQ Table le Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 89.72 89.72 71.31 71.31 2 3.37 93.09 15.87 87.18 3 3.77 96.86 2.61 89.79 4 1.85 98.71 1.10 90.89 5 0.39 99.10 3.35 94.24 6 0.40 99.50 1.18 95.42 7 0.33 99.83 0.82 96.24 8 0.04 99.86 1.54 97.78 9 0.08 99.94 0.24 98.02 1H NMR Prediction Model for Density (g/ml) of RCC Feed

Mod

del

SLIDE 8

50 100 150 200

0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1

0.1 Variable LV 1 (89.72%) Variables/Loadings Plot for Binned Data.csv

50 100 150 200

1
0.8
0.6
0.4
0.2

0.2 0.4 Variable LV 4 (1.85%) Variables/Loadings Plot for Binned Data.csv 50 100 150 200

0.3
0.2
0.1

0.1 0.2 0.3 0.4 Variable LV 9 (0.08%) Variables/Loadings Plot for Binned Data.csv

Examples of Latent Variable Loadings for LV1, LV4, and LV9

SLIDE 9

5 10 15 20 0.5 1 1.5 2 Hotelling T^2 (99.95%) Q Residuals (0.05%) 0.2 0.4 0.6 0.8

3
2
1

1 2 3 Leverage Y Stdnt Residual 1 10 15 20 25 30 35 40 10 15 20 25 30 35 Y Measured 1 Y CV Predicted 1

50

50 100

40
20

20 40 Scores on LV 1 (88.67%) Scores on LV 2 (5.66%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 06-Oct-2014 15:54:029.09 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 47 by 248 Included: [ 2-10 12-14 16-35 38-46 48-53 ] [ 1-248 ] Preprocessing: Mean Center Y-block: API Gravity - 54 Samples.xlsx 47 by 1 Included: [ 2-10 12-14 16-35 38-46 48-53 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 9

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.51 deg RMSECV: 0.77 deg Bias: -1.42109e-014 CV Bias: 0.0102836 R^2 Cal: 0.98737 R^2 CV: 0.971025 SSQ Table le Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 88.67 88.67 72.12 72.12 2 5.66 94.33 11.86 83.98 3 2.69 97.02 5.02 89.00 4 1.88 98.90 1.12 90.12 5 0.41 99.30 2.74 92.86 6 0.39 99.70 1.84 94.70 7 0.17 99.87 1.57 96.27 8 0.03 99.91 2.06 98.33 9 0.04 99.95 0.41 98.74 1H NMR Prediction Model for API Gravity of RCC Feed

SLIDE 10

5 10 15 0.02 0.04 0.06 0.08 Hotelling T^2 (99.35%) Q Residuals (0.65%) 0.1 0.2 0.3 0.4

4
3
2
1

1 2 3 Leverage Y Stdnt Residual 1 0.05 0.1 0.15 0.2 0.25 0.05 0.1 0.15 0.2 0.25 Y Measured 1 Y CV Predicted 1

4
2

2 4

1
0.5

0.5 1 Scores on LV 1 (93.95%) Scores on LV 2 (3.41%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 07-Oct-2014 23:45:055.39 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 51 by 97 Included: [ 2-35 38-54 ] [ 1 153-248 ] Preprocessing: Mean Center Y-block: Fa_C13.xlsx 51 by 1 Included: [ 2-35 38-54 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 3

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.0067 RMSECV: 0.0074 Bias: -2.77556e-017 CV Bias: 0.0002 R^2 Cal: 0.975 R^2 CV: 0.970 SSQ SSQ Tab able Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 93.95 93.95 88.54 88.54 2 3.41 97.36 5.84 94.38 3 2.00 99.35 3.15 97.53 Analysis - PLS 3 LVs - 54 Samples - Binned Data - X-Block.xlsx, Fa_C13.xlsx

1H NMR Prediction Model for Fraction Carbon Aromaticity of RCC Feed

Primary Test Method: 13C NMR

SLIDE 11

5 10 15 0.02 0.04 0.06 0.08 Hotelling T^2 (99.63%) Q Residuals (0.37%) 0.1 0.2 0.3 0.4

3
2
1

1 2 3 Leverage Y Stdnt Residual 1 1 2 3 4 1 2 3 4 Y Measured 1 Y CV Predicted 1

4
2

2 4

2
1.5
1
0.5

0.5 1 Scores on LV 1 (93.93%) Scores on LV 2 (4.35%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 08-Oct-2014 00:26:012.79 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 48 by 88 Included: [ 2-14 16-30 32-35 38-51 53-54 ] [ 1 162-248 ] Preprocessing: Mean Center Y-block: Sulfur - 54 Samples.xlsx 48 by 1 Included: [ 2-14 16-30 32-35 38-51 53-54 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 5

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.17 wt% RMSECV: 0.26 wt% Bias: -1.33227e-015 CV Bias: 0.004 R^2 Cal: 0.958 R^2 CV: 0.902 SSQ SSQ Tab able Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 93.93 93.93 55.11 55.11 2 4.35 98.28 17.67 72.78 3 1.11 99.39 12.16 84.94 4 0.17 99.56 7.02 91.96 5 0.08 99.63 3.86 95.82

1H NMR Prediction Model for Sulfur Content (Wt%) of RCC Feed

Primary Method - XRF

SLIDE 12

5 10 15 0.05 0.1 0.15 0.2 Hotelling T^2 (98.96%) Q Residuals (1.04%) 0.1 0.2 0.3 0.4

4
2

2 4 Leverage Y Stdnt Residual 1 5 10 15 20 25 5 10 15 20 25 Y Measured 1 Y CV Predicted 1

4
2

2 4

0.5

0.5 Scores on LV 1 (93.82%) Scores on LV 2 (0.85%)

1H NMR Prediction Model for Total Aromatic Content (Wt%) of RCC Feed

Primary Method – HPLC-UV-DAD

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 16-Sep-2014 00:39:028.36 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 49 by 73 Included: [ 1 3-33 35 38-51 53-54 ] [ 1-2 159-229 ] Preprocessing: Mean Center Y-block: Total Aromatics - 54 Samples.xlsx 49 by 1 Included: [ 1 3-33 35 38-51 53-54 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 3

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.80 wt% RMSECV: 0.88 wt% Bias: 3.55271e-015 CV Bias: -0.02 R^2 Cal: 0.949 R^2 CV: 0.937 SSQ SSQ Tab able Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 93.82 93.82 88.47 88.47 2 0.85 94.67 6.38 94.85 3 4.29 98.96 0.07 94.92

SLIDE 13

2 4 6 8 10 0.05 0.1 0.15 0.2 Hotelling T^2 (99.16%) Q Residuals (0.84%) 0.05 0.1 0.15 0.2

4
2

2 4 Leverage Y Stdnt Residual 1 1 2 3 4 5 6 1 2 3 4 5 6 Y Measured 1 Y CV Predicted 1

4
2

2 4

1
0.5

0.5 1 Scores on LV 1 (95.08%) Scores on LV 2 (4.07%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 16-Sep-2014 00:33:023.38 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 47 by 25 Included: [ 1 3-4 6-8 10-16 18-33 35 38-54 ] [ 1 169-192 ] Preprocessing: Mean Center Y-block: MonoAromatics - 54 Samples.xlsx 47 by 1 Included: [ 1 3-4 6-8 10-16 18-33 35 38-54 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 2

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.22 wt% RMSECV: 0.23 wt% Bias: 0 CV Bias: 0.0006 R^2 Cal: 0.946 R^2 CV: 0.937 SSQ SSQ Tab able Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 95.08 95.08 92.99 92.99 2 4.07 99.16 1.63 94.62

1H NMR Prediction Model for Mono-Aromatic Content (Wt%) of RCC Feed

Primary Method – HPLC-UV-DAD

SLIDE 14

5 10 15 20 0.01 0.02 0.03 0.04 0.05 0.06 Hotelling T^2 (99.61%) Q Residuals (0.39%) 0.2 0.4 0.6 0.8

4
2

2 4 Leverage Y Stdnt Residual 1 1 1.5 2 2.5 3 3.5 4 1 1.5 2 2.5 3 3.5 4 Y Measured 1 Y CV Predicted 1

4
2

2 4

1
0.5

0.5 1 1.5 2 Scores on LV 1 (93.62%) Scores on LV 2 (4.95%)

1H NMR Prediction Model for Di-Aromatic Content (Wt%) of RCC Feed

Primary Method – HPLC-UV-DAD

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 06-Oct-2014 18:44:054.01 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 44 by 92 Included: [ 1 3-4 6-8 10 12-33 35 38-45 47-50 53-54 ] [ 1 158-248 ] Preprocessing: Mean Center Y-block: DiAromatics.xlsx 44 by 1 Included: [ 1 3-4 6-8 10 12-33 35 38-45 47-50 53-54 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 4

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.11 wt% RMSECV: 0.13 wt% Bias: 0 CV Bias: 0.004 R^2 Cal: 0.973 R^2 CV: 0.959 SSQ SSQ Tab able Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 93.62 93.62 90.30 90.30 2 4.95 98.57 3.24 93.54 3 0.57 99.15 2.91 96.45 4 0.47 99.61 0.86 97.31

SLIDE 15

5 10 15 0.05 0.1 0.15 0.2 Hotelling T^2 (98.59%) Q Residuals (1.41%) 0.1 0.2 0.3 0.4

3
2
1

1 2 3 Leverage Y Stdnt Residual 1 1 2 3 4 5 1 2 3 4 5 Y Measured 1 Y CV Predicted 1

4
2

2 4

1
0.5

0.5 1 1.5 2 Scores on LV 1 (95.12%) Scores on LV 2 (3.47%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 08-Oct-2014 00:51:00.601 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 41 by 60 Included: [ 3 6-8 10-14 16-18 20-23 25-31 33 35 38-51 53-54 ] Preprocessing: Mean Center Y-block: TriAromatics.xlsx 41 by 1 Included: [ 3 6-8 10-14 16-18 20-23 25-31 33 35 38-51 53-54 ] Preprocessing: Mean Center

Num. LVs: 2

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.22 wt% RMSECV: 0.24 wt% Bias: -4.44089e-016 CV Bias: -0.002 R^2 Cal: 0.939 R^2 CV: 0.930 SSQ SSQ Tab able Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 95.12 95.12 93.21 93.21 2 3.47 98.59 0.75 93.96

1H NMR Prediction Model for Tri-Aromatic Content (Wt%) of RCC Feed

Primary Method – HPLC-UV-DAD

SLIDE 16

10 20 30 40 0.5 1 1.5 2 Hotelling T^2 (99.97%) Q Residuals (0.03%) 0.2 0.4 0.6 0.8

3
2
1

1 2 3 Leverage Y Stdnt Residual 1 2 4 6 8 10 5 10 15 Y Measured 1 Y CV Predicted 1

100
50

50 100

40
20

20 40 Scores on LV 1 (87.46%) Scores on LV 2 (6.13%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 08-Oct-2014 00:54:036.52 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 48 by 248 Included: [ 1-4 6-14 16-23 25-33 35-36 38-51 53-54 ] [ 1-248 ] Preprocessing: Mean Center Y-block: tetraaromatc.xlsx 48 by 1 Included: [ 1-4 6-14 16-23 25-33 35-36 38-51 53-54 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 11

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.31 wt% RMSECV: 0.55 wt% Bias: 1.77636e-014 CV Bias: 0.0146361 R^2 Cal: 0.968 R^2 CV: 0.906

SSQ Table le Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 87.46 87.46 11.16 11.16 2 6.13 93.59 27.76 38.92 3 2.72 96.31 18.22 57.14 4 1.84 98.15 3.50 60.64 5 1.06 99.21 4.21 64.85 6 0.43 99.64 8.14 72.99 7 0.08 99.72 15.70 88.68 8 0.15 99.87 2.83 91.51 9 0.02 99.89 4.12 95.64 10 0.06 99.95 0.33 95.97 11 0.02 99.97 0.79 96.76 1H NMR Prediction Model for Tetra-Aromatic Content (Wt%) of RCC Feed

Primary Method – HPLC-UV-DAD

SLIDE 17

5 10 15 20 25 0.5 1 1.5 Hotelling T^2 (99.97%) Q Residuals (0.03%) 0.2 0.4 0.6 0.8

3
2
1

1 2 3 Leverage Y Stdnt Residual 1 20 25 30 35 40 20 25 30 35 40 Y Measured 1 Y CV Predicted 1

100
50

50

20
10

10 20 Scores on LV 1 (89.69%) Scores on LV 2 (3.70%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 06-Oct-2014 22:11:050.31 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 46 by 248 Included: [ 2-10 12-14 16-35 38-43 45-46 48-53 ] [ 1-248 ] Preprocessing: Mean Center Y-block: V50.xlsx 46 by 1 Included: [ 2-10 12-14 16-35 38-43 45-46 48-53 ] [ 1 ] Preprocessing: Mean Center

Num. LVs: 10

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.47 RMSECV: 0.74 Bias: 1.42109e-014 CV Bias: 0.003 R^2 Cal: 0.975 R^2 CV: 0.938 SSQ SSQ Tab able

Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 89.69 89.69 36.91 36.91 2 3.70 93.39 17.59 54.50 3 3.52 96.91 6.59 61.08 4 1.90 98.81 3.43 64.51 5 0.51 99.32 12.24 76.74 6 0.35 99.67 9.48 86.23 7 0.19 99.86 2.61 88.84 8 0.04 99.91 5.79 94.63 9 0.04 99.95 1.29 95.92 10 0.02 99.97 1.61 97.53 1H NMR Prediction Model for Vicosity Blending Index (V50) of RCC Feed

Primary Method – Kinematic Viscosity Measurement – Refutas Equation * * Baird, C. T. (1989). Guide to petroleum product blending. Austin (TX): HPI Consultants, Inc

SLIDE 18

5 10 15 20 25 30 0.05 0.1 0.15 0.2 Hotelling T^2 (99.99%) Q Residuals (0.01%) 0.2 0.4 0.6 0.8

4
3
2
1

1 2 3 Leverage Y Stdnt Residual 1 40 50 60 70 80 90 100 40 50 60 70 80 90 100 Y Measured 1 Y CV Predicted 1

100
50

50

40
30
20
10

10 20 Scores on LV 1 (89.95%) Scores on LV 2 (5.68%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 08-Oct-2014 01:19:024.97 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 43 by 104 Included: [ 3-10 12-13 15-21 23-24 26 28-30 32-35 38-46 48-54 ] Preprocessing: Mean Center Y-block: Dist_425_Plus.xlsx 43 by 1 Preprocessing: Mean Center

Num. LVs: 8

Cross validation: venetian blinds w/ 7 splits RMSEC: 2.62 wt% RMSECV: 3.70 wt% Bias: 1.42109e-014 CV Bias: 0.12 R^2 Cal: 0.936 R^2 CV: 0.874 SSQ SSQ Tab able Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 89.95 89.95 52.11 52.11 2 5.68 95.63 9.75 61.86 3 2.86 98.49 1.32 63.18 4 1.23 99.72 2.33 65.51 5 0.09 99.81 16.22 81.73 6 0.17 99.98 4.74 86.47 7 0.01 99.99 4.84 91.31 8 0.00 99.99 2.30 93.62

1H NMR Prediction Model for RCC Feed Distillation – 470C_Plus

Primary Method – Distillation –

SLIDE 19

10 20 30 40 0.5 1 1.5 Hotelling T^2 (99.95%) Q Residuals (0.05%) 0.2 0.4 0.6 0.8

3
2
1

1 2 3 Leverage Y Stdnt Residual 1 5 10 15

5

5 10 15 Y Measured 1 Y CV Predicted 1

100
50

50

20
10

10 20 Scores on LV 1 (84.76%) Scores on LV 2 (4.60%)

Mod

del

Linear regression model using Partial Least Squares calculated with the SIMPLS algorithm Developed 08-Oct-2014 01:33:021.37 Author: John@JCEPNA X-block: 54 Samples - Binned Data - X-Block.xlsx 43 by 135 Preprocessing: Mean Center Y-block: MCRT - 54 Samples.xlsx 43 by 1 Preprocessing: Mean Center

Num. LVs: 9

Cross validation: venetian blinds w/ 7 splits RMSEC: 0.37 wt% RMSECV: 0.68 wt% Bias: 1.77636e-015 CV Bias: 0.046 R^2 Cal: 0.981 R^2 CV: 0.934 SSQ SSQ Tab able Percent Variance Captured by Regression Model

----X-Block-----
----Y-Block-----

Comp This Total This Total

------ -------
------ -------

1 94.91 94.91 16.59 16.59 2 2.10 97.01 47.14 63.73 3 1.85 98.87 17.64 81.37 4 0.41 99.27 5.92 87.28 5 0.38 99.65 4.31 91.60 6 0.19 99.84 1.15 92.75 7 0.04 99.88 4.06 96.81 8 0.04 99.92 0.86 97.67 9 0.02 99.94 0.39 98.06

1H NMR Prediction Model for RCC Feed Micro Carbon Residue Test (MCRT)

Primary Method – MCRT - ASTM D4530

SLIDE 20

Parameter R2 (CV) SECV Density 0.954 0.0052 g/ml API Gravity 0.971 0.77 deg

13C Carbon Aromaticity Fraction

0.97 0.0074 Sulfur 0.902 0.26 wt% Total Aromatics 0.937 0.88 wt% Mono-Aromatics 0.937 0.23 wt% Di-Aromatics 0.959 0.13 wt% Tri-Aromatics 0.93 0.24 wt% Tetra-Aromatics 0.906 0.55 wt% Viscosity Blending Index – V50 0.938 0.74 Distillation - 470 Deg_Plus 0.874 3.7 wt% Micro Carbon Residue Test 0.934 0.68 wt% Summary of Partial Least-Squares Regression Analysis of 1H NMR and VGO Quality Parameters

SLIDE 21

Gasoline – 1H NMR Spectral Variability Aromatics Olefins Oxygenate Alpha – CH3 CH3 CH2 CH Gasoline Quality Parameters Correlated to NMR RON, MON, R+M/2, D86 Distillation, Flash Point, Aromatics, Benzene, Olefins, RVP, Sulfur Saturates, FIA, SARA, Oxygenates

Validation Calibration

Parameter Variance SECV R2 Factors SEC R2 Density 99.06 0.0024 g/ml 0.9984 5 0.0020 g/ml 0.9991 Aromatics 98.79 0.77 Vol% 0.9994 5 0.64 Vol% 0.9996 Benzene 99.33 0.19 Vol% 0.9744 5 0.12 Vol% 0.9925 Olefins 99.73 1.41 Vol% 0.9938 6 1.15 Vol% 0.9967 MTBE 99.03 0.27 Vol% 0.9986 5 0.10 Vol% 0.9999 IBP 99.96 5.12 oC 0.9636 6 4.26 oC 0.9783 T10 99.49 4.22 oC 0.9905 6 3.59 oC 0.9942 T50 99.07 2.13 oC 0.9967 5 1.61 oC 0.9983 T90 99.81 2.15 oC 0.9982 7 1.41 oC 0.9989 FBP 99.57 4.11 oC 0.9896 6 3.28 oC 0.9946 RON 99.67 0.31 0.9968 8 0.22 0.9989 MON 99.79 0.35 0.9971 8 0.25 0.9989 RVP 99.87 0.026 bar 0.9956 9 0.014 bar 0.9990 Sulfur 99.69 61.7 ppm 0.9820 6 39.0 ppm 0.9949

SLIDE 22