SPECTROSCOPY BASED CRUDE ASSAYS FOR LABORATORY, AT-LINE, AND ON-LINE APPLICATIONS

Transcription

1 SPECTSCPY BASED CUDE ASSAYS F LABATY, AT-LINE, AND N-LINE APPLICATINS Paul J. Giammatteo and John C. Edwards Process NM Associates, LLC Danbury, Connecticut

2 Traditional True Boiling Point which accurately measures product yields Water Content API Gravity Metals and Inorganic Content (nickel, vanadium, sulfur, sodium, chlorine) Corrosivity Spectroscopic Full distillation yields (True boiling point vol% and wt%) by NM Water content (Microwave water cut and NM) API Gravity (NM) Carbon distribution (aromatics, olefins, naphthenics, paraffins) (NM) Metals (nickel, vanadium) (ES) Acidity/Corrosivity (ES / Microwave)

3 NM Nuclear Magnetic esonance» Time-Domain/Low esolution» High esolution ES Electron Spin esonance FTI Fourier Transform Infra ed with Diamond AT» Mid-infra-red region ( 6) LIBS - Laser Induced Breakdown Spectroscopy 1. Can be applied in a variety of laboratory, at-line, and/or online configurations. 2. ffer rapid, cost effective and simultaneous multi-parameter analyses. 3. Measurement of both bulk and trace properties. 4. Ideal for use in combination with software planning tools.

4 Crude il Chemistry: Alkanes, cycloalkanes (naphthenes), aromatics, polycyclic aromatics, S-containing compounds, etc. Gasoline: branched alkanes Diesel: linear alkanes

5 Polycyclic Aromatics and Asphaltene Precursors

6 il efining, one of the most complicated manufacturing operations in all industry, taking extremely diverse feed stocks to make products of consistent quality worldwide.

7 2 NM Spectra Showing Spectral Variability of 6 Crudes CH 2 Gasolines CH 3 CH All Diesels: (Data date range 1999 to present) Atmospheric Crude Straight un Vacuum Crude Straight un CH 2 CH 3 Coker Diesel CH 3 CH 3 Diesel Hydro Cracker Feed Diesel Hydro Cracker Product Mild, Moderate and Severely Hydro Treated Diesel Blend Header Diesel CH 3 FCC Diesel H H H H H ppm

8 2 NM Spectra Showing Spectral Variability of 6 Crudes CH 2 CH 3 CH CH 3 CH ppm

9 Table 1: Initial Study Crude ils - 23 to 28 Alaskan North Slope Forties Nile Blend Amna Fosterton dudu Arab Heavy Gos Mix ily ock Arab Medium Grand Isle Qatar Marine Arab X-light High Island Qua Iboe Bach Ho Kerapu ang Dong Badak Khafji as Garib Basra Kiame atawi (Neutral Zone) Basra Light Kirkuk oss Bay Marchand Kuwait uby Bay Marchand Lago Medio Schiehallion BCF-17 Lavan Blend Shell Delta Belayim (Egypt) Light Louisiana Sweet South Arne Brent Liverpool BlendStag Cheleken Loyal Syrian Light Cuisiana Lu Feng Tantawan Cupiago Mandji Tapis Blend Cupiagua Marib light Tengiz Curlew Mars Torcados (Nigeria) Didon Masa Upper Zakum Dubai Masila Wafra (Neutral Zone) Dulang Maui Condensate Walio Dulang Maya West Texas Intermediate Duri Minas Xi Xiang Es Sahara Mudi YME Flotta Nemba Zafiro Blend

10 TBP Weight Percent Cut Point (Actual versus NM Predicted) Partial Least Squares results for Table 1 Crude oils. Graphs are for (clockwise) deg C, 2 deg C, 7 deg C, and, deg C. Correlation Coefficients ( 2 ) for each plot are greater than.96 8 Actual Crude TBP Curves 7 19 NM Predicted TBP at C (Wt%) Actual NM Predicted TBP at 2C (Wt%) 2 7 Actual Cumulative Weight Loss (Wt%) Temperature (C) Kuwait Marib Light Nemba GS Mix Dulang Bacho NM Predicted TBP at C (Wt%) Actual NM Predicted TBP at 7C (Wt%) Actual Cumulative Weight Loss (Wt%) NM Predicted TBP Curves for Crudes Temperature (C)

11 pna--9.tdf,9 (² = ) pna--9.tdf, (² = ) Predicted Concentration ( F17 C1 ) Predicted Concentration ( F C1 ) Normal-Paraffins of -9C Cut Iso-Paraffins of -9C Cut Actual Concentration ( C1 ) Actual Concentration ( C1 ) pna--9.tdf,14 (² = ) 14 pna--9.tdf,19 (² = ) Predicted Concentration ( F14 C1 ) Naphthenes of -9C Cut Predicted Concentration ( F17 C1 ) Aromatics of -9C Cut Actual Concentration ( C1 ) Actual Concentration ( C1 ) density-api-s.tdf,12 (² = ) con res-asphaltenes.tdf,2 (² = ) Predicted Concentration ( F16 C1 ) API Gravity Predicted Concentration ( F16 C1 ) Carbon esidue Actual Concentration ( C1 ) Actual Concentration ( C1 )

12 Water in Crude By Process 1 H NM 9 ² =.9982 rmsd=.11 Wt% Wt% Predicted Water Content (Wt%) Actual Water Content (Wt%)

13 Crude Analysis

14 Lagomar/Basrah 71/29 GP(Lagotreco/DZ/Boscan) 9/4/1 GP(Lagotreco/DZ/Boscan/Slops) 9/4/1/ GP(Lagotreco/DZ) 9/ GP(Lagotreco/riente/DZ/Slops) 68/23/4/ GP(Lagotreco/riente/DZ) 92/3/ GP(Lagotreco/riente/mesa/DZ) 74/14/7/ GP(Lagotreco/riente/mesa/DZ/slops) 1/11/28// GP(Lagotreco/LagotrcoMediano/riente/mesa/DZ/Slops) 37/2/24/28/4/ GP(DZ/MESA/Lagotreco/LagotrcoMediano/rientae/Slops) /27/42/6/12/3 GP = General Purpose Crude. Slops = refinery slop re-blends of unknown composition and quality.

15 Crude Adjustment Blue line represents the predicted curve using only an off-line LP program based on historical crude properties of the blend. W T % Y i e l d 6 LP econ NM 3 Magenta line is the actual yield curve for the same crude 2 blend obtained using the reconciliation function of the optimizer. The triangles are the feed forward NM predicted yield curve demonstrating the NM accuracy compared to the LP Cut Point Deg C

16 Crude API 8/2-9/6/1 :3:38 :16:27 6:29:22 21:42:2 12::32 4:8:17 19:21:2 :33:2 1:46:49 18:49:43 ::13 1:3:48 16:1:3 8:12:22 :48:7 21:7:14 12:2:33 23:32:13 14:3:9 6:14:16 Crude Switch Tank Switch Feed change causing CDU upset.

17 6 MHz NM of on-line and laboratory analyzed crude samples. ed Spectrum is ASTM 93 (Cushing) run in PNA lab

18 .7.6 lefins in crude are typically from the blending of cracked stocks found in refinery slops, condensate, etc, or from thermal cracking in enhanced recovery processes distinct water peaks are observed, some centered at 74 and most centered at 78. Water peak position is indicator of polarity of crude matrix. lefins from coker diesel for reference

19 .8 Aromatic and Water egion Expansion ed Spectrum is ASTM n-line spectra showing two distinct water peaks. Crude spectra are after pre-heaters and desalters. Water position is indicator of crude polarity

20 Predicted Concentration ( F8 C1 ) Predicted Concentration ( F8 C1 ) SimDis T combined crude simdis t(f) 629.tdf,6 (² = ) SimDis T combined crude simdis t (f) 629.tdf, (² = ) Actual Concentration ( C1 ) Predicted Concentration ( F8 C1 ) Predicted Concentration ( F8 C1 ) combined crude simdis t3(f) 629.tdf, (² =.93916) SimDis T3 SimDis T combined crude simdis t7 (f) 629.tdf,6 (² =.9648) Actual Concentration ( C1 ) Actual Concentration ( C1 ) Actual Concentration ( C1 )

21 combined crude simdis t9 (f) 629.tdf,7 (² = ) 12 Predicted Concentration ( F7 C1 ) 1 9 SimDis T Fraction 1 (Vol%) Actual Concentration ( C1 ) day API Gravity Same model - two crude feeds Fraction 2 (Vol%)

22 Comparison of Bitumen And Precipitated Asphaltenes Bitumen 8 Asphaltenes ring

23 An electron spin resonance (ES) spectrometer detects the concentration and composition of chemical species with unpaired electrons that are present in a sample. Typical chemical species are free radicals and many transition metal ions, depending on their valence state. The sample is loaded into a high frequency resonant cavity in a magnetic field. Free radicals irradiated with microwave radiation will undergo transitions at a characteristic frequency Sample Number Location 1 Vasconia, Magdalena Basin, Columbia 39 2 Merey, Eastern Venezuela Basin, Venezuela 33 3 riente, riente Basin, Ecuador 6 4 Qua Iboe, Agbada Formation, Niger Delta 2 Basrah Light, Zubair Zone, Iraq 29 Vanadium Content (ppm) riginal Assay

24 Kero Blnk BasLt Merey riente QuaIboe Vasconia

25 Kero Blnk BasLt Merey riente QuaIboe Vasconia

26 SAMPLING and TESTING All Diesels: (Data date range 1999 to present) CH 2 Atmospheric Crude Straight un Vacuum Crude Straight un CH 3 Coker Diesel FCC Diesel Diesel Hydro Cracker Feed Diesel Hydro Cracker Product Mild, Moderate and Severely Hydro Treated Diesel Blend Header Diesel CH 3 H H H H H

27 Crude Diesel SimDis D86 Numerous factors (especially aromatics) are responsible for deviations between distillation methods. SimDis (actual) Corr D86 Actual D % 9% 9 % 96% 97% 9 8% 99% FB D S T T T 9 9 L C G T T T 9 HCU CH T T T 9 D S T T T 9 L A D T T T 9 B/MA T T T 9 CAB DS T T T 9 7% 8 % 6% 6 % 7% 4% % % 3% 3 % % 2% IB 1 % 2 % 3% 4 % % % % 2 %

28 Numerous factors (especially aromatics) are responsible for deviations between distillation methods. SimDis (actual) Corr D86 Actual D DSL T T T9 9 LCG T T T9 HCU CHG T T T9 DSL T T T9 LAD T T T9 B/MAD T T T9 CAB DSL T T T9

29 ISSUES Traditional and Spectroscopic Sampling, Sampling, Sampling epresentative Sample Time and Money Accuracy Value of the Information Questions?