U.S. patent application number 14/260388 was filed with the patent office on 2014-09-18 for direct match spectrographic determination of fuel properties.
The applicant listed for this patent is TESORO REFINING AND MARKETING COMPANY. Invention is credited to Daniel C. Mertens.
Application Number | 20140278141 14/260388 |
Document ID | / |
Family ID | 47142442 |
Filed Date | 2014-09-18 |
United States Patent
Application |
20140278141 |
Kind Code |
A1 |
Mertens; Daniel C. |
September 18, 2014 |
DIRECT MATCH SPECTROGRAPHIC DETERMINATION OF FUEL PROPERTIES
Abstract
A method and apparatus for deriving a refinery product property
value based on data produced from a globally-calibrated
spectrographic analyzer and data from a non-spectrographic
analyzer.
Inventors: |
Mertens; Daniel C.; (San
Antonio, TX) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
TESORO REFINING AND MARKETING COMPANY |
San Antonio |
TX |
US |
|
|
Family ID: |
47142442 |
Appl. No.: |
14/260388 |
Filed: |
April 24, 2014 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
13493749 |
Jun 11, 2012 |
8735820 |
|
|
14260388 |
|
|
|
|
13084500 |
Apr 11, 2011 |
8481942 |
|
|
13493749 |
|
|
|
|
61322755 |
Apr 9, 2010 |
|
|
|
Current U.S.
Class: |
702/22 |
Current CPC
Class: |
G01N 21/3577 20130101;
G01N 21/359 20130101; G01N 21/65 20130101; G01J 3/28 20130101; G01N
33/28 20130101; G01N 33/00 20130101; G01N 21/274 20130101; G01N
2021/3595 20130101; G01N 33/2823 20130101 |
Class at
Publication: |
702/22 |
International
Class: |
G01N 33/00 20060101
G01N033/00 |
Claims
1-27. (canceled)
28. An apparatus for direct match comparison, the apparatus
comprising a spectrographic analyzer capable of receiving a
property value of a first refinery product using a
non-spectrographic test; determining a property value of the first
refinery product using spectrographically determined data stored in
a spectral library; determining a property value of a second
refinery product; determining a difference between the
spectrographically-determined property values of the first refinery
product and the second refinery product; and adding the difference
to the non-spectrographically-determined property value of the
first refinery product to derive a property value for the second
refinery product.
29. A system for direct match comparison, the system comprising: a
spectrographic analyzer capable of determining a property value of
a test refinery product; and a computer programmed to receive a
property value of a reference refinery product, the property value
determined using a non-spectrographic test; receive a property
values of the reference refinery product from data stored in a
spectral library; receive said property value of the test refinery
product determined by the spectrographic analyzer; determine a
difference between the spectrographically-determined property
values of the test refinery product and the reference refinery
product; and add the difference to the
non-spectrographically-determined property value of the reference
refinery product to derive a property value for the test refinery
product.
30. An apparatus for direct match comparison, the apparatus
comprising a computer programmed (A) to receive (1) a property
value of a first refinery product, the property value determined
using a non-spectrographic test, (2) a
spectrographically-determined property value of the first refinery
product determined from data stored in a spectral library, and (3)
a property value of a second refinery product determined using a
spectrographic analyzer; (B) to determine a difference between the
spectrographically-determined property values of the first refinery
product and the second refinery product; and (C) add the difference
to the non-spectrographically-determined property value of the
first refinery product to derive a property value for the second
refinery product.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] This application claims priority to U.S. Nonprovisional
application Ser. No. 13/084,500 filed Apr. 11, 2011, which in turn
claims priority to U.S. Provisional Patent Application 61/322,755
entitled "Direct Match Spectrographic Determination of Fuel
Properties" filed Apr. 9, 2010, each of which is hereby entirely
incorporated herein by reference.
FIELD
[0002] The disclosed method and apparatus generally relate to
spectrographic determination of fuel properties.
BACKGROUND
[0003] Hydrocarbons, such as crude oil, may be refined to produce
various products, such as jet fuel, gasoline, diesel fuel,
paraffins, kerosene, naphtha, lubricating oils, asphalts, fuel oil,
and liquefied petroleum gases (LPGs) such as propane and butane.
Refining generally refers to a group of processes that treat and
chemically change hydrocarbons. The refining process for crude oil
generally breaks apart the heavier, or more dense, hydrocarbon
chains of the crude oil at various pressures and temperatures to
produce lighter, or less dense, hydrocarbon fractions. The refining
process may result in finished products, such as diesel fuel, and
may result in intermediate products, such as fuel oil, that require
further processing to produce a finished product. For example, a
fuel oil may be refined further through conversion to change its
chemical components in a way suitable for use in blending
gasoline.
[0004] Crude oil may be made up of hundreds of chemical components.
Such components may include, for example, alkanes, aromatics,
olefins, isomers, and napthenes. Some of those components are
heavier than others, and the conversion process may make some
components heavier, or may make them lighter, or may change them in
some other way to result in certain properties. The conversion
process may include, for example, distillation, coking,
hydrocracking, fluidized catalytic cracking (FCC), alkylation,
de-sulfurization, reforming and isomerization. The conversion
process may also rely on catalysts, such as platinum, and other
process variables, such as temperature and pressure to effect
conversion. Thus, product components may change during the refining
process. The refining process may also include blending various
intermediate products and additives to create a finished product.
Additives may include, for example, ethers and alcohols.
[0005] Each product may have various chemical and physical
properties that particularly relate to that product's usefulness.
Some of those properties may be related to particular components
such as benzene. Other properties may be related to the product
overall, such as specific gravity. For example, increasing octane
may be useful in reducing engine "knocking" when burned, and
reducing sulfur may result in lower levels of harmful sulfur
dioxide, a combustion by-product.
[0006] Petroleum products may be produced to specification. Thus, a
refinery in a certain geographic location may develop a product
matrix to meet a certain specification based on the types of crude
oil that it processes. For example, a retail gasoline station in
the northwestern United States may request gasoline blended to have
a certain grade, such as regular, or be of a certain type, such as
RFG (ReFormulated Gasoline), and be suitable for a certain season,
such as winter. A refiner may then develop a product matrix for its
location that may characterize the requested grade, type and
season. To develop a certain gasoline blend, a refinery may mix
various intermediate products, such as those from a vacuum tower,
hydrocracker, FCC unit and/or alkylation unit. Different matrices
may be provided for other products, and those matrices may vary in
complexity. For example, a product matrix for diesel fuel may
characterize different components than a gasoline product
matrix.
[0007] Additionally, petroleum products must comply with the
environmental and other regulations of the state in which the
product will be sold and used. For example, an environmental
regulation may require reduced sulfur in the finished product.
[0008] Thus, prior to sending a product to a retailer for sale, a
refinery may test or analyze the finished product and/or
intermediate products (such as, for example, blendstock for
oxygenate blending, or BOB) to determine the properties of the
products and to verify those properties against the product
specification and regulatory standards. Industry standards, such as
ASTM standards, may in some cases define how such tests or analyses
are conducted, and often, regulatory agencies will rely on industry
standard test procedures to certify a product for sale or use.
[0009] Spectrographic analyzers having location- and
matrix-specific calibrations may be used by refineries to determine
product properties. However, because such calibrations may be
location- and matrix-specific, methods using spectrographic
analyzers may not comply with the requirements of existing industry
standard methods of certifying a fuel product as having specific
properties. Regulatory agencies may thus refuse to accept test
results based on spectrographic analysis as valid for product
certification. In addition, location- and matrix-specific
calibrations may make it difficult to compare data between
different instruments, sampling or processing techniques, and
laboratories, which may, for example, be physically located at the
same or different location. For example, a refiner may have
multiple refineries in different geographic locations, each
processing a different crude oil and using different instruments.
Therefore, there is a need for a method and apparatus for
standardizing product property values determined by spectrographic
analysis.
SUMMARY
[0010] A method for deriving a refinery product property value, the
method comprising spectrographically analyzing a test product to
obtain test product spectral data; searching a spectral data
library comprising reference spectral data associated with at least
one reference product; selecting reference spectral data from a
reference product among said at least one reference product;
wherein a part of the selected reference spectral data resembles a
part of said test product spectral data; determining a property
value of the test product using a globally-calibrated
spectrographic analyzer; determining a property value for said
reference product using data stored in said spectral data library
and global calibration data; determining a difference between the
spectrographically-determined property values of the test product
and the reference product; and adding the difference to a
non-spectrographically-determined property value of the reference
product to derive a certified property value for the test
product.
[0011] A method for deriving a refinery product property value, the
method comprising determining a property value of a first refinery
product using a non-spectrographic test; determining a property
value of the first refinery product using a first
globally-calibrated spectrographic analyzer; determining a property
value of a second refinery product using a second
globally-calibrated spectrographic analyzer; determining a
difference between the spectrographically-determined property
values of the first refinery product and the second refinery
product; adding the difference to the
non-spectrographically-determined property value of the first
refinery product to derive a property value for the second refinery
product.
[0012] An apparatus for direct match comparison, the apparatus
comprising a spectrographic analyzer capable of receiving a
property value of a first refinery product using a
non-spectrographic test; determining a property value of the first
refinery product using a first globally-calibrated spectrographic
analyzer; determining a property value of a second refinery
product; determining a difference between the
spectrographically-determined property values of the first refinery
product and the second refinery product; and adding the difference
to the non-spectrographically-determined property value of the
first refinery product to derive a property value for the second
refinery product.
[0013] A system for direct match comparison, the system comprising
a (1) spectrographic analyzer capable of determining a property
value of a first refinery product, and determining a property value
of a second refinery product; and (2) a computer programmed to
receive a property value of the first refinery product, the
property value determined using a non-spectrographic test; receive
the property values of the first and second refinery products
determined by the spectrographic analyzer; determine a difference
between the spectrographically-determined property values of the
first refinery product and the second refinery product; and add the
difference to the non-spectrographically-determined property value
of the first refinery product to derive a property value for the
second refinery product.
[0014] An apparatus for direct match comparison, the apparatus
comprising a computer programmed (A) to receive (1) a property
value of a first refinery product, the property value determined
using a non-spectrographic test, (2) a property value of the first
refinery product determined using a spectrographic analyzer, and
(3) a property value of a second refinery product determined using
the spectrographic analyzer; (B) to determine a difference between
the spectrographically-determined property values of the first
refinery product and the second refinery product; and (C) add the
difference to the non-spectrographically-determined property value
of the first refinery product to derive a property value for the
second refinery product.
BRIEF DESCRIPTION OF THE DRAWINGS
[0015] FIG. 1 depicts one embodiment of a method for producing
global calibration data.
[0016] FIG. 2 depicts one embodiment of a method for determining a
reference value.
[0017] FIG. 3 depicts one embodiment of a method for direct match
comparison of a test product to a reference product.
[0018] FIG. 4 depicts one embodiment of a method of searching a
database library and for direct match comparison of a test product
to a reference product.
[0019] FIG. 5 depicts one embodiment of a method for direct match
comparison of a prototype product to a reference product.
[0020] FIG. 6 depicts one embodiment of a method for direct match
comparison of a test product to a reference product or prototype
product.
[0021] FIG. 7 depicts one embodiment of a system for performing a
direct match comparison.
DETAILED DESCRIPTION
[0022] Spectrographic analyzers, or spectrometers, may include, for
example, those associated with the following wavelengths or
techniques: Near Infrared (NIR), Mid Infrared (MIR), Near and Mid
(full range) Infrared (IR), Fourier Transform Near Infrared
(FTNIR), Fourier Transform Mid Infrared (FTMIR), Fourier Transform
Near and Mid (full range) Infrared (FTIR), Nuclear Magnetic
Resonance (NMR) and Raman. By way of nonlimiting example, such an
analyzer may be a stand-alone instrument suitable for batch
testing, such as those produced by ABB Bomem, and by Analect or may
be an on-line instrument (i.e., connected to fuel production
equipment) suitable for in-stream testing, such as those produced
by ABB Bomem, and Analect. Such an analyzer may be implemented in
hardware and/or software.
[0023] Spectrographic analyzers may be used to quantify properties
of, for example, lubricants (nominally lubricating oils), spark
ignition fuels (nominally gasoline), turbine fuels (nominally jet
fuel), and distillate fuels (nominally diesel fuel and home heating
oils.). Use of such analyzers may achieve an improved precision
(e.g., less test variation) when compared to other available
industry techniques, such as engine testing.
[0024] A refinery may calibrate the spectrographic analyzer to the
product properties. For example, the properties of fuels may
include, but are not limited, to those shown in Table 1:
TABLE-US-00001 TABLE 1 Spark Ignition Turbine Compression
Distillate Fuel (e.g. Fuel Ignition Fuel Fuels Property Gasoline)
(e.g. Jet) (e.g., Diesel) (e.g. Htr Oil) RON X MON X RVP X
T.sub.v/l=20 X Specific X X X X Gravity Aromatics X X X Polynuclear
X Aromatics Olefins X Benzene X Oxygen X Ethanol X Distillation X X
X X Flash X X X Viscosity X X X Analine Point X Cetane Number X
"RON" refers to Research Octane Number. "MON" refers to Motor
Octane Number. "RVP" refers to Reid Vapor Pressure.
T.sub.(v/l)=20refers to the temperature at which the vapor to
liquid ratio equals 20. Distillation properties may be obtained
with respect to IBP, T10, T30, T50, T70, T90, EP, E200 and E300,
for example.
[0025] To calibrate a spectrometer to a product's properties and
property values, a refinery may take multiple samples of particular
product over time, and send a first portion of each sample to a
laboratory for primary (non-spectrographic) testing, e.g., engine
testing, according to ASTM standards, such as ASTM D2700, for
testing motor octane number (MON) of spark-ignition engine fuels.
The refinery may run a second portion of each sample through the
spectrographic analyzer to generate spectral data, and compare the
primary test data to the spectral data in order to correlate
product properties and property values to the spectral data. For
example, a refinery may determine the relationship between an
engine test MON value of a sample against that sample's wavelength
absorption data. Each comparison generates calibration data. A
refinery may aggregate the data and use statistical tools to
calibrate the spectrographic analyzer to specific product property
values. For example, a refinery may use a multivariate regression
analysis, such as that specified in ASTM E1655, to develop a
calibration curve from the calibration data. After calibration, a
refinery may use the spectrographic analyzer to analyze the
properties of a product sample and predict property values without
having to also analyze the sample using primary or
non-spectrographic testing.
[0026] Although using a spectrographic analyzer may provide a more
precise result, product matrix biases, or the non-linearity of
various product properties, can adversely affect the accuracy of
the property value measurements. These adverse biases may be
addressed by creating location-specific and matrix-specific
calibrations. For a gasoline blend, for example, a product matrix
may be dependent upon the refinery location, gasoline grade (e.g.,
regular/premium), gasoline type (e.g., RFG (ReFormulated Gasoline),
RBOB (Reformulated Blendstock for Oxygenate Blending),
Conventional, CARB (California Air Resources Board), CARBOB
(California Air Resources Blendstock for Oxygenate Blending)), and
season (e.g., winter, summer). A refinery could thus have several
separate calibrations that would cover all or some of its specific
types, grades and seasons of gasoline, and these calibrations would
generally only include data from that specific refinery (or its
laboratory), and may not include data from other refineries (or
outside laboratories). A refinery may similarly develop
calibrations for other intermediate and finished products, as
well.
[0027] Creating and using location-specific and matrix-specific
calibrations may raise a number of challenges, such as a general
lack of confidence that after the calibration is created, any
property value will be measured accurately (extrapolation). Another
challenge may be that the predicted value may be difficult to
replicate at refineries or laboratories in other locations that
process and test other products. A further challenge may be a lack
of acceptance of property values for certification by an industry
group or regulatory agency.
[0028] Furthermore, a product may be developed to have certain
property values, and may be certified on the basis of certain
property values. Certain property values such as research octane
number (RON) and motor octane number (MON) may, for example,
indicate how a fuel may behave in operation, such as when burned in
an engine. However, some refineries may optimize the RON and MON
(and other property values) in different ways, including, for
example, by addition of different ratios of blending components or
by processing different crude oils or other intermediate products.
Thus, different products may have the same property values, but may
show significantly different chemical signatures when analyzed by a
spectrographic analyzer.
[0029] One way to resolve location- and matrix-specific biases may
be to generate global calibration values. A global calibration may
be based on spectral and analytical data from a variety of
location- and matrix-specific calibrations. "Global" may refer to
worldwide refinery product, or as much refinery product as may be
available from participating refineries. In some embodiments,
global calibration data may be produced from a data set of
participating users, and may include, for example, data produced
from a range of blend components, fuel properties or a combination
of both. The global calibration values may be based on location-
and matrix-specific data for a particular product, such as
gasoline, produced by a variety of refineries, or may be based on a
variety of products from one or more refineries. Global calibration
data for gasoline, for example, may include data from a variety of
locations and from a variety of product matrices. A group of
refineries may aggregate their location- and matrix-specific
calibrations, and use a statistical tool to develop global
calibration values. For example, a multivariate regression
analysis, such as that described in ASTM E1655, may be used to
develop global calibration values for various properties.
[0030] In some embodiments, spectral data that is included in
global calibration may be associated with a region of the spectrum
where one or more species that may be present in a fuel are
vibrationally active, e.g., where such species may experience one
or more molecular oscillations. In some embodiments, spectral data
that is included in global calibration may be associated with a
region of the spectrum where the data is related to different
refineries or where the data is related to different seasonal
varieties of a type of fuel. In some embodiments, a spectral region
may be selected because including data from such a region may
affect the spread of data included in global calibration, because
inclusion of data from such regions affects the local curvature of
global calibration data, or both.
[0031] FIG. 1 depicts an exemplary embodiment of a method 10 of
developing global calibration data for a refinery product
comprising fuel. In step 2, two or more refineries may produce one
or more fuel products, such as gasoline. Of course, other types of
refinery products, such as diesel and jet fuel may be produced.
Each product may have its own product matrix. For example, a
refinery at one geographic location may have a gasoline product
matrix characterized by grade, type and season. A refinery at a
different geographic location may have a different gasoline product
matrix characterized by grade, type and season.
[0032] In step 4, each refinery may test each product using
non-spectrographic (e.g., engine testing, distillate testing, etc)
and spectrographic testing (e.g., using a spectrographic analyzer)
in order to determine product values and spectral data pertaining
to each product. Of course, outside laboratories may test the
refinery products for a refinery, or provide testing in addition to
a refinery's own testing. Data may be determined based on each
product, or on one or more matrix variables, or on one or more
product properties (such as those listed in Table 2), or on the
entire output of a refinery, or on any other desired basis suitable
for globally calibrating a spectrographic analyzer for analyzing
refinery product. Spectrographic data included for the purpose of
global calibration may include data from the entire spectral region
collected during spectrographic testing or may include data, e.g.,
such as absorption, from a more narrow spectral region.
[0033] In step 6, each refinery (or laboratory) may provide the
data collected in step 4 for aggregation. For example, each
refinery (or laboratory) may send its data to a spectrographic
analyzer manufacturer, or to an industry group for aggregation.
[0034] In step 8, the aggregated product data may be statistically
treated, as discussed above, in order to develop global calibration
data for each refinery product. Thereafter, the global calibration
data may be provided to one or more of the participating
refineries, as well as to non-participating refineries and
laboratories.
[0035] A manufacturer of the spectrographic analyzer may develop
global calibration values, and provide such values along with its
spectrographic analyzers. A manufacturer may periodically update
the global calibration values as participating refineries and
laboratories create or refine their location- and matrix-specific
calibrations and provide that data to the manufacturer. Of course,
other parties may develop global calibration values, such as
industry groups, vendors, vendor groups or laboratories. Thus, each
laboratory or refinery may have one or more globally-calibrated
spectrographic analyzers, and those global calibrations may be the
same for all refineries and laboratories using a particular
manufacturer's equipment.
[0036] However, using global calibration values to determine
property values of a particular product at a particular refinery
requires compensating for location- and matrix-specific biases. To
do so, a refinery may develop a reference product and use a direct
match technique to correct such biases.
[0037] By way of example, a reference product comprising a fuel may
be used. A reference fuel may be collected as a sample. The
reference fuel collected may have a composition including matrix
components that is related to the refinery from which it is
collected. The reference fuel may then undergo primary testing
using a non-spectrographic instrument or analyzer to determine the
reference fuel's property values. For example, a reference fuel may
be captured at a specific refinery by carefully collecting and
storing a volume of fuel representative of that refinery's spark
ignition fuel or blendstock, such as gasoline, of a specific grade,
type and season. The octane values of the sample may then be tested
by engine testing to determine a robust value for each of the
properties of interest to the refinery. The octane values
determined by engine testing may be statistically treated to
develop a reference value for each octane value. A statistical
treatment may involve, for example, using the GESD technique in
accordance with ASTM Research Report D02-1481, and assess all
non-rejected data for normality at a 1% significance level in
accordance with the Anderson-Darling technique in Standard Practice
D6299.
[0038] Of course, as noted above, other types of primary testing
may be used to determine a reference fuel's property values. Other
types of products, such as diesel fuels, may undergo primary
testing based on other standards, such as distillate testing under
ASTM D86. Other primary tests may rely on, for example, RVPe
testing under ASTM D5191, and benzene testing by gas chromatography
under ASTM D3606. Other non-spectrographic techniques suitable for
primary testing of fuel property values may be used.
[0039] FIG. 2 illustrates an exemplary method 20 of determining the
reference value of a fuel. In a step 22 of FIG. 2, a refinery may
collect a sample of fuel to be used as a reference. In step 24, a
refinery may perform primary or non-spectrographic testing on the
reference sample to obtain product data. Alternatively, as
discussed in more detail below, a refinery may send out the sample
for testing by an outside laboratory, if a more rigorous approach
is desired. In step 26, a refinery may statistically treat the
product data, as discussed above, to produce a reference value for
at least one product property.
[0040] In some embodiments, a refinery may choose among three
levels of rigor in primary testing to determine reference values:
laboratory reference values, consensus reference values and
semi-consensus reference values. Of course, reference values may be
developed in other ways. Those levels may provide a trade-off
between assurance of accuracy vs. time and expense. One level of
rigor may simply be a laboratory reference value determined by
tests run at the refiner's own laboratory. Generally, a laboratory
reference value may be determined from two or more tests done at
the refinery's own laboratory. For this level, the refinery may run
20 tests on portions of the same sample of reference fuel, and
determine the robust mean average of the test data. In other
embodiments, a robust laboratory reference value may be determined
by taking the robust mean average of data from at least 16
individual tests run within that refinery's laboratory, using as
many different analyzers and technicians as possible.
[0041] To determine the accuracy of laboratory reference values, a
refinery may use inter-laboratory cross-check data to verify that
the reference analyzer(s) are unbiased against other laboratories.
This may be accomplished by statistical quality control (SQC)
charting the difference between the reference sample analyzer test
data and outside laboratory test data to assure accuracy. Robust
SQC charting may be used to determine, at the 2-sided 5% level,
whether there is any significant bias between the reference sample
analyzer data and outside laboratory data.
[0042] To calculate a laboratory reference value, a refinery may,
for example, assess all candidate data (e.g., excluding statistical
outliers) using the GESD technique in accordance with ASTM Research
Report D02-1481, and assess all candidate data for normality at a
1% significance level in accordance with the Anderson-Darling
technique in Standard Practice D6299. The calculation may involve
tabulation of all the non-rejected data, determining the mean
average of the non-rejected data, and determining the standard
deviation of all the sample data. The reference value and standard
deviation may be specified to a useful numerical precision. For
example, the following values may be reported with the following
precision:
[0043] RON: XX.XXX
[0044] MON: XX.XXX
[0045] S.G at 60 F: X.XXXXX
[0046] RVP (ptot): XX.XXX
[0047] T.sub.v/1=20: XXX.X
[0048] Aromatics" XX.X
[0049] Olefins: XX.X
[0050] Benzene: X.XX
[0051] Ethanol vol %: XX.X
[0052] Oxygen wt %: X.XX
[0053] Distillation points: XXX.X
[0054] Anti-knock index (AKI) or Road Octane may be determined by
the mean average of RON and MON, and then rounded down to the
nearest 0.5 octane number.
[0055] Other reference values may be further calculated based on
the determined value. For example, ASTM, EPA or CARB RVPe may be
calculated from the PTOT value defined in ASTM D5191.
[0056] A higher level of rigor may involve developing a consensus
value. Generally, a consensus value may be determined from two or
more tests done at one or more outside laboratories. A consensus
value may be determined by sending portions of the reference sample
to, e.g., 20 different laboratories for analysis. A robust mean
average of the test data from those laboratories may be used to
develop a consensus value for each property of interest.
[0057] For example, a consensus value for a reference fuel may be
the mean average of data from at least 16 individual tests, run by
20 different analyzers and 20 technicians, using a specified
reference test method. For example, a reference test method may be
specified for aromatics in an EPA-regulated gasoline as ASTM D5769
(GC mass spec), and for aromatics in a CARB-regulated gasoline as
ASTM D5580.
[0058] To calculate a consensus value, a refinery may, for example,
assess all candidate data (e.g., excluding statistical outliers)
using the GESD technique in accordance with ASTM Research Report
D02-1481, and assess all candidate data for normality at a 1%
significance level in accordance with the Anderson-Darling
technique in Standard Practice D6299. The calculation may involve
tabulation of all the non-rejected data, determining the mean
average of the non-rejected data, determining the standard
deviation of all the sample data, and determining if two standard
deviations (2.sigma.) is within published reproducibility (R) of
the reference test method. The consensus value and standard
deviation may be specified to a useful numerical precision.
[0059] An intermediate level of rigor may involve developing a
semi-consensus value by combining laboratory reference test data
and consensus test data. Generally, a semi-consensus value may be
determined from data produced from the refinery's own laboratory
and from at least one outside laboratory. For example, a
semi-consensus value may involve determining the numeric average of
the robust mean average of data from 20 tests at the refiner's
laboratory and the robust mean average of data from five tests run
at five outside laboratories. The use of the robust mean of data
from five outside laboratories may provide for any small
corrections for accuracy.
[0060] To determine the accuracy of semi-consensus values, a
refinery may use inter-laboratory cross-check data to verify that
the reference analyzer(s) are unbiased against other laboratories.
This may be accomplished by statistical quality control (SQC)
charting the difference between the reference analyzer test data
and outside laboratory test data to assure accuracy. Robust SQC
charting may be used to determine, at the 2-sided 5% level, whether
there is any significant bias between the reference analyzer data
and outside laboratory data.
[0061] To calculate a semi-consensus value, a refinery may assess
all candidate data (e.g., excluding statistical outliers) using the
GESD technique in accordance with ASTM Research Report D02-1481,
and assess all candidate data for normality at a 1% significance
level in accordance with the Anderson-Darling technique in Standard
Practice D6299. The calculation may involve tabulation of all the
non-rejected data from the reference laboratory tests, determining
the mean average of the non-rejected data, and determining the
standard deviation of all the sample data. The calculation may
involve tabulation of all the non-rejected data from the outside
laboratory tests, determining the mean average of the non-rejected
data, and determining the standard deviation of all the sample
data. The calculation may involve determining the mean average of
the reference laboratory non-rejected data and the outside
laboratory non-rejected data. The reference value and standard
deviation may be specified to a useful numerical precision.
[0062] The reference product sample remaining after primary testing
may be carefully preserved to maintain the integrity of its
chemical composition. The preserved reference sample, or portions
thereof, may be later used in connection with testing a
chemically-similar product having unknown property values. A
refinery may thus take care in collecting and storing samples. For
example, manual samples for laboratory batch analysis may be
collected in accordance with ASTM Standard Practices D4057 and
D5842. Samples may be taken in accordance with any other suitable
technique. Stream samples for on-line analysis may be collected in
accordance with ASTM Standard Practice D4177. Sample fuels may be
collected and stored in opaque containers, such as an amber or dark
brown glass bottle, to minimize exposure to UV emissions from
sources such as sunlight or fluorescent lamps. Alternatively,
samples may be stored in clear bottles contained within boxes that
prevent light exposure.
[0063] To test an unknown product against the reference product, a
refinery may take a sample of a test product that is deemed to be
chemically similar to the reference product. The test sample may be
spectrographically tested together with a portion of the preserved
reference sample under tightly controlled conditions.
[0064] Regarding chemical similarity, in some embodiments, a test
product may be deemed chemically similar to a reference product
based on chemical analysis. For example, chemical analysis may
include characterization of the test and reference product using a
spectrographic analyzer or with some other analytical instrument
capable of quantitative chemical analysis. Chemically similar
products may, for example, have a similar distribution of
hydrocarbons such as octane, iso-octane, heptanes, other straight
or branched chain hydrocarbons, or combinations thereof. It should
be appreciated that comparison of the reference product and test
product may in some embodiments be performed one or more times.
More generally, the comparison of the reference product and test
product may be performed a fewer number of times than analysis of
test samples. Therefore, comparison of reference and test samples
may economically include comparison of those samples using a
greater number of techniques than associated with testing of test
product. In some embodiments, the performance of the reference and
test product in a test engine or some other non-spectrographic
instrument may be used in the comparison.
[0065] In some embodiments, chemical similarity may be determined
with references to particular properties, or by comparing spectral
data. Spectral data may be used to verify or support a belief that
a test product suspected to be similar to reference sample, e.g.,
that a test product produced in a similar way to reference sample
is similar to the reference sample. Alternatively, spectral data
may be used to select a particular reference sample from among a
group of possible reference samples as suitable for comparison with
a given test product. For example, a database library may be
created that includes one or more spectra for each of a group of
reference products. The reference product or products for which
spectra are included in a library may comprise all available
reference products for which spectral data are available or a group
of products that are related because they share a certain
characteristic, e.g., production origin at a certain refinery,
seasonal grade or type. To select a certain reference product, the
spectra of an unknown product may be collected and an operator may
execute a search of the database to find a particular reference
spectrum that most closely resembles the spectrum of the unknown
product. Following the selection or verification of a reference
fuel, the reference fuel may, in some embodiments and as further
described below, be tested sequentially together with the test
product.
[0066] In comparing particular properties, for example, chemical
similarity may be defined within a certain range. For example,
chemical similarity based on octane may be defined as no more than
.+-.2 octane number difference. In some embodiments, a test fuel
may be deemed chemically similar to a reference fuel based on the
following criteria:
[0067] RON within .+-.2 Octane Number,
[0068] MON within .+-.2 Octane Number,
[0069] RVPe within .+-.20.7 kPa (3.0 psia),
[0070] Specific Gravity.+-.0.1,
[0071] Aromatics.+-.10 vol %,
[0072] Olefins.+-.10 vol %,
[0073] T10.+-.5.6.degree. C. (20.0.degree. F.),
[0074] T50.+-.5.6.degree. C. (20.0.degree. F.),
[0075] T90.+-.16.7.degree. C. (40.0.degree. F.),
[0076] EP.+-.16.7.degree. C. (40.0.degree. F.),
[0077] Wt % Oxygen.+-.1.0 wt %
[0078] Vol % Ethanol.+-.4.0 vol %.
[0079] Generally, a refinery may have prior knowledge of a
product's chemical properties as a result of specifying the
refining (including blending) process. A refinery may know, for
example, that a premium gasoline is much different from regular
gasoline because of the absence or presence of certain components,
such as alcohols. In other words, different products may generally
be different in composition, not just different in degree.
[0080] A chemically-similar test product and reference product may
be tested together by spectrographic analysis. In the example of
fuel, to certify a chemically-similar test fuel against a reference
fuel, a direct match comparison technique may be used. For each
certification, a portion of the preserved reference fuel sample may
be obtained from a sealed container or from a piston type or water
displacement system. Both the reference fuel and test fuel may be
chilled in the same way prior to testing. For example, the
temperature may be greater than 0.degree. C. (32.degree. F.) but
not exceed 10.degree. C. (50.degree. F.). All comparisons may be
accomplished using a single globally-calibrated spectrometer.
Alternatively, different globally-calibrated spectrometers may be
used. The reference fuel and test fuel may be tested in sequence
for one or more cycles, e.g., 10 cycles. However, in some
embodiments, the results of a previously executed spectrographic
analysis, such as may be provided, for example, by a highly precise
and accurate FTIR spectrometer, may be used instead of testing the
reference fuel together, e.g., in sequence, with running the test
fuel. For example, an operator may use a stored spectrum for a
reference product that has previously been measured with high
precision and accuracy.
[0081] For batch analysis of a test fuel in the laboratory, for
example, the method may comprise (1) obtaining a sample of
reference fuel (or prototype fuel, as discussed in more detail
below) of the same type, grade and season as the fuel to be tested;
(2) chilling both the reference fuel and the test fuel in the same
manner and for the same length of time; (3) determining the
reference fuel properties using the globally-calibrated
spectrometer; (4) determining the test fuel properties using the
globally-calibrated spectrometer; and (5) calculating the delta
between the test fuel and the reference fuel for each property:
.DELTA..sub.i=TI.sub.i-RI.sub.i (Equation 1)
Where:
[0082] .DELTA..sub.i=difference between the
spectrographically-determined property values of the test fuel and
the reference fuel [0083] TI.sub.i=spectrographically-determined
property value of the test fuel (text index) [0084]
RI.sub.i=spectrographically-determined property value of the
reference fuel (reference index) [0085] i=a property, such as MON
or RON
[0086] The method may further comprise calculating the derived
property of the test fuel (T.sub.i):
T.sub.i=.DELTA..sub.i+R.sub.i (Equation 2)
Where:
[0087] T.sub.i=derived property value of the test fuel for
certification [0088] R.sub.i=property value of the reference fuel
determined by primary testing (e.g., consensus value, laboratory
reference value or semi-consensus value)
[0089] Thus, the direct match comparison technique may comprise
adding the delta (.DELTA..sub.i) to the property value of the
reference fuel determined by primary testing (R.sub.i) to establish
the property value of the test fuel that may be used for
certification. In various embodiments, R.sub.i may comprise the
consensus value, the laboratory reference value or semi-consensus
value of the reference fuel.
[0090] FIG. 3 illustrates an exemplary embodiment of a method 30
for direct match comparison of a test fuel against a
chemically-similar reference fuel. In a step 32, a sample of a test
fuel for which analysis of a property value is desired may be
collected. In a step 34, a reference fuel may be selected for use
in a direct match analysis of the test fuel. The test fuel may be
chemically similar to the reference fuel. Chemical similarity may
be determined in the manner discussed in more detail above.
[0091] In a step 36, a refinery's laboratory may run a
spectrographic test of the test fuel and the reference fuel using a
globally-calibrated spectrometer. For example, as shown in Table 2,
the RON for the test fuel may be determined spectroscopically to be
92.00 and the RON for the reference fuel may be determined to be
spectroscopically 92.10. The MON for the test fuel may be
determined to be 82.30 and the MON for the reference fuel may be
spectroscopically determined to be 81.15. In step 38, the
differences, or deltas, between the spectrographically-determined
test and reference values may be calculated as in Table 2.
TABLE-US-00002 TABLE 2 Reference Reference Test Fuel Fuel Fuel
Spectroscopic Spectroscopic Consensus TEST Determination
Determination Delta Value Test Result RON 92.00 92.10 -0.10 92.21
92.11 MON 82.30 81.15 1.15 82.35 83.50
[0092] Consensus values of the reference fuel may be known because
the consensus values may have been previously measured, such as in
a manner as described in above. Of course, laboratory reference or
semi-consensus values may be used as well. For example, a consensus
value for RON of the reference fuel may be 92.21 and a consensus
value for MON of the reference fuel may be 82.35. In step 40, the
RON and MON value deltas may be added to the reference consensus
RON and MON values, respectively, to derive the RON and MON values
of the test fuel suitable for certification.
[0093] For analysis of a test fuel in an on-line analyzer, for
example, the method may comprise (1) lining up the correct
reference (or prototype fuel, as discussed in more detail below) of
the same type, grade and season as the fuel to be blended; (2) when
the blend starts, testing the blender sample three times, then the
reference sample once, then the blender sample three more times,
all using one or more globally-calibrated spectrographic analyzers;
(3) determining the average of each property of the six blender
samples; (4) determine the delta between the averaged test sample
property values and the reference fuel property values (using
Equation 1); and (5) adding the delta to the consensus property
value (or laboratory reference or semi-consensus value) of the
reference fuel (using Equation 2) to derive a property value of the
test fuel for certification.
[0094] The method may further comprise latching the determined
property value T.sub.i for the next cycle after completing the test
fuel analysis, determining the volume of gasoline that passed
through the blend header since the last reported value, and
integrating the property values into the total blended volume. In
some embodiments, after a number of cycles, the sample test cell in
the on-line analyzer may be chemically cleaned and checked with a
reagent, such as toluene.
[0095] In some embodiments, the Simulated Knock Intensity (SKI) of
the RON value of the test fuel may be used as a continuous
indicator of the health of the on-line system. Of course, other
suitable properties could be used as health indicators. The SKI may
start out as the value 50 at the beginning of the blend. Depending
on the condition of the sample cell and other external factors, the
SKI value may be monitored through the course of the blend. Any
significant deviation from the value 50 may be used to create an
alarm to check the system.
[0096] Thus, for example, on the first cycle, the Simulated Knock
Intensity (SKI) value may be set to 50, and the initial determined
RON of the test fuel noted. On subsequent cycles, the SKI of the
test fuel may be determined as follows:
SKI.sub.r=50+round((IDO.sub.r-DO.sub.r)*12,2)
where:
[0097] SKI.sub.r=current simulated knock intensity of test fuel
[0098] IDO.sub.r=initial determined RON of test fuel on startup
[0099] CDO.sub.r=current determined RON of test fuel
[0100] In some embodiments, acceptance criteria for on-line
analysis may include ensuring that the analyzer was on-line for 95%
of the blend, the RON test fuel SKIs during the blend were within
.+-.10 SKI of the startup SKI for 90% of the blend, and desired
property value conditions were met for 90% of the blend. If an
on-line analyzers is able to demonstrate significant consistency
throughout a blend, as demonstrated by SQC charting, the frequency
of testing the test fuel may be reduced as long as the analyzer
remains demonstrably in control.
[0101] In some embodiments, a product with unknown properties may
be spectrographically tested, and spectral characteristics of the
product may be compared to a library of spectral data of cataloged
reference products. As used herein, the term "cataloged reference
product" may refer to a one or more reference products for which
spectra have been recorded and for which at least a portion of the
collected spectral data may be stored, e.g., in a spectral database
or library. If spectral characteristics of the unknown product and
one or more catalogued reference product are found to be suitably
alike, a reference product may be deemed to be chemically similar
to the product with unknown properties. The reference product may
thus be selected for testing in a direct match method for
determining a product property value. Selection of a reference
product based on a relatively high degree of chemical similarity
may serve to minimize the effects of matrix variation and tend to
minimize differences between the test and reference sample which
may, e.g., minimize sources of errors associated with non-linear
calibrations.
[0102] In some embodiments, a product with unknown properties may
be spectrographically tested and compared to a library of spectral
data associated with various catalogued reference products. A
program may search the spectral library and identify spectral data
within the library that most closely matches the spectrum of the
product with unknown properties. If data in the spectral library
meets certain criteria, such as, e.g., precision and/or similarity
with the test data, the stored data may be used for a reference
index. Thus, in some embodiments, a spectrographically-determined
property value of a reference fuel (reference index) may be based
on data stored within or calculated from a spectral library. The
use of a reference index that is selected or derived from
pre-determined data stored in a spectral library may be
appropriate, e.g., where a reference product provides upon analysis
with a certain technique, such as, e.g., Fourier transform infrared
spectroscopy, highly consistent spectral data. In some embodiments,
the entire function of the reference fuel, including, e.g.,
calculations involving spectrographic and primary testing property
values, may be performed using entirely previously obtained
data.
[0103] For analyzing a test fuel using a direct match approach and
for searching of a spectral library, e.g., the method may comprise
(1) obtaining a sample of test product; (2) analyzing the test fuel
using spectroscopy; (3) searching of a spectral library; (4)
selecting an appropriate catalogued reference product; (5)
determining reference fuel properties using spectral data included
in the library; (6) determining test fuel properties; (7)
calculating the delta between the test fuel and the reference fuel
for each desired property; and (8) calculating the derived
properties of the test fuel. The method may utilize equations 1 and
2 described previously. However, RI.sub.i (the reference index) may
be derived spectrographically from previously collected data, e.g.,
data stored in a spectral library or linked to the spectral
library.
.DELTA..sub.i=TI.sub.i-RI.sub.i (Equation 1)
T.sub.i=.DELTA..sub.i+R.sub.i (Equation 2)
[0104] Thus, a direct match comparison technique may comprise
adding the delta (.DELTA..sub.1) to the property value for the
reference fuel determined from primary testing (R.sub.i) to
establish the property value of the test fuel that may be used for
certification. In various embodiments, R.sub.i may comprise the
consensus value, the laboratory reference value or semi-consensus
value of the reference fuel.
[0105] FIG. 4 illustrates an exemplary embodiment of a method 42
for direct match comparison of a test product using data stored in
a spectral library. In a step 44, a sample of a test fuel for which
analysis of a property value is desired may be collected. In a step
46, the test fuel (or a portion of the test fuel) may be
spectrographically analyzed. Spectrographic analysis may, e.g.,
comprise collection of absorption data in an infrared region of the
spectrum, such as NIR, MIR, or the full IR range, and may use an
instrument, such as, e.g., a FTIR spectrometer, that provides
highly consistent data. In a step 48, spectrographic data for the
test fuel may be compared to spectral data stored in a reference
library, i.e., a search of the library database may be performed to
identify any reference spectra that resemble the spectrum for the
test fuel.
[0106] In some embodiments, a complete search of the spectra of all
included reference fuels in the spectral library may be performed.
For example, a spectral library may include data from all available
reference fuels and a search of the entire spectral library may be
performed. In some embodiments, a spectral library may comprise or
be partitioned to allow for the search of only spectral data
derived from reference fuels of a given product or from testing a
group of reference fuels which match a preset classification, such
as, e.g., reference fuels of a specific seasonal type, grade,
refinery (or limited number of related refineries), or reference
fuels classified by one or more other criteria. In some
embodiments, it may be known that a given reference fuel or
reference fuel of a given type may be suitable for use in the
reference fuel function and searching a spectral library may find
the particular spectrum from a list of recorded spectra stored for
the specific reference fuel that is the best match to product test
data.
[0107] In some embodiments, a spectral library may be searched and
a metric related to similarity between test fuel spectral data and
data in the spectral library may be established. A degree of
similarity may, e.g., be provided to an operator, and maybe used as
criteria for selection of the method 42. Also, an operator may opt
to perform additional tests, such as, e.g., tests associated with
supporting analytical techniques, should the degree of similarity
between test data and catalogued reference data be low. Any of
various techniques, such as, e.g., correlation based approaches
and/or least squared methodologies, may be used to assess the
similarity of test data and library data.
[0108] In some embodiments, it may be known that a particular fuel
may include one or more chemical additives (such as an alcohol or
an ether) or it may be known that additives were added to a given
fuel within certain concentration ranges, and only the spectra of
catalogued reference fuels that are known to include that additive
or that are specified to include that additive in a given
concentration range may be searched. Alternatively, all catalogued
reference fuels may be searched; however, spectra that are
consistent, such as qualitatively or quantitatively, with the
presence of a certain chemical additive may be given more
statistical weight than other spectra. For example, spectral
regions associated with the presence of an alcohol, such as
wavenumbers of about 3200 cm.sup.-1 to about 3550 cm.sup.-1, or
spectral regions associated with the presence of an ether, such as
wavenumbers of about 1000 cm.sup.-1 to about 1300 cm.sup.-1, may,
in some embodiments, be searched or weighted.
[0109] Searching a spectral library and/or evaluating the
similarity of test data and library data may involve selecting all
spectral frequencies for which data is available or the selection
of a subset of the available frequency data, such as, e.g., the
fingerprint region of the IR spectrum. In some embodiments,
searching a library database may involve an initial assessment of
the similarity of test spectral data and library data using a first
level of rigor and for certain portions of data, e.g., those that
with the first level of rigor meet a threshold level of similarity,
using other levels of rigor, such as to obtain a more accurate
assessment of similarity. Thus, it should be understood that
searching of a library database and assessment of similarity
between spectral data may be run simultaneously or sequentially as
may, e.g., be appropriate for optimizing speed and/or accuracy. In
some embodiments, spectral frequencies used for searching and/or
evaluation of similarity may comprise substantially the same
regions of the spectrum as may be included in global calibration
data. In some embodiments, spectral frequencies may be selected for
searching and/or evaluation of similarity because those frequencies
correspond to molecular oscillations associated with a fuel
additive or component. In addition, as noted above, data associated
with the frequency of a certain molecular oscillation may be
weighted of greater importance in assessing similarity.
[0110] In some embodiments, analysis techniques, including, e.g.,
multivariate testing methodologies such as principal component
analysis (PCA), may be used to estimate an expected concentration
range for one or more components detected in a fuel. An operator
may review a concentration range and may, e.g., use such analysis
in assessing the validity of the search.
[0111] Data catalogued in a spectral library may, for any given
reference fuel, comprise the entirety of spectral data for a given
wavelength region, data from regions of a spectrum used for global
calibration data, data from other spectral regions or combinations
thereof. To that point, frequency bands useful for global
calibration and for searching a database may be the same or
different. In some embodiments, a particular reference spectrum may
be selected from a library because selection of that particular
reference spectrum may minimize a difference between the
spectrographically-determined property values of the test fuel and
the reference fuel. Catalogued data may include one or more
spectra, such as from duplicate testing of a reference fuel sample,
and may, e.g., include information related to the precision of
measurement of a reference sample, including data for measurements
repeated on a given day or for measurements repeated over a longer
time period, such as historical data. Data included in a spectral
library may further include or be linked to associated data for a
given catalogued reference fuel. Data linked to a spectral library
may include a consensus, laboratory reference or semi-consensus
value for any desired property (e.g., RON, MON, RVP, Specific
Gravity, etc.) and may, in some embodiments, further include
additional data, such as, e.g., the date a given spectrum was
collected, associated tool data (such as serial number of a
specific tool used for data collection), date a sample was
obtained, method of optimization for a given property (such as
addition of certain additives), grade, seasonal type, refinery
location, characteristics of associated crude oil, any established
concentration ranges for a certain fuel component(s), other
observations concerning a reference fuel, and supporting analytical
data. Supporting analytical data may include, e.g., other chemical
analysis, such as, e.g., Raman spectroscopy, NMR, GC-Mass Spec, or
other analytical techniques.
[0112] In a step 50, a refinery's laboratory may run a
spectrographic test of the test fuel using a globally-calibrated
spectrometer. Spectrographic testing of a sample may incorporate or
use data collected in step 46 of method 42, and it may be
convenient to do so. A test fuel may be measured in a
spectrographic analyzer any number of times as may be desired to
generate sufficient statistics. Using global-calibration data a
spectrometer may determine the value of desired properties. For
example, the RON for the test fuel may be determined
spectroscopically to be about 92.25 and the MON for the test fuel
may be 81.20. A property value for the reference fuel may be
determined based on global calibration data and spectral data
stored within the spectral library. Thus, the method 42 may use a
spectrographically-determined property value that is established
from testing that was previously performed. Therefore, method 42
may, e.g., alleviate some concerns associated with storage and/or
re-running valuable reference fuel concurrently with test fuel
analysis. By way of example, the RON for a reference fuel may be
determined from database information and global calibration data to
be about 93.00 and the MON may be about 81.35.
[0113] In a step 52, the difference, or delta, between the
spectrographically-determined test and reference value may be
calculated. For example, using the model data above the delta for
the RON may be -0.05 (92.25-93.00=-0.05). Using the model data
above the delta for the MON may be -0.15 (81.20-81.35=-0.15). In a
step 54, the consensus value of the reference fuel (or laboratory
reference or semi-consensus value) for any desired property may be
added to the deltas to derive a certified property value. Consensus
values, laboratory references or semi-consensus values may, e.g.,
have been previously measured as described above. For example, a
consensus value for the RON may be 92.01 and a consensus value for
the MON may be 82.85. Data for the model example described above
and in relation to method 42 is shown in Table 3.
TABLE-US-00003 TABLE 3 Reference Test Fuel Reference Fuel Fuel
Spectroscopic (Determination Consensus Test TEST Determination from
database data) Delta Value Result RON 92.25 93.00 -0.05 92.01 91.96
MON 81.20 81.35 -0.15 82.85 82.70
[0114] If relatively small amounts of the reference product volume
are used for testing (e.g., 1 gallon/month for laboratory
analysis), the reference product may be used in connection with
analyzing test product. However, if the usage for testing is larger
(e.g. used for an on-line analyzer), the reference product may be
used to certify prototype products for use in testing test
products, which in turn may be used in connection with product
testing. In addition, for some embodiments, a reference product may
be analyzed and spectral data may be stored for future use in
analyzing test product. In some embodiments, a prototype product
and a reference product may be collected from the same refinery or
a refinery of similar design using comparable procedures. In some
embodiments, a prototype and a reference product may be collected
from the same refinery or a refinery of similar design and may use
a source of crude oil collected from the same geographic location.
In some embodiments, a prototype product and a reference product
may be processed in the same refinery, or a refinery of similar
design, and may be processed using one or more of the same or
similar additives. In some embodiments, a prototype product and a
reference product may be processed in at least one common process
or unit associated with the same refinery.
[0115] Generally, the reference product and prototype product may
be collected in the same manner and preserved to maintain their
vapor pressure and chemical stability. The collection protocol may
be the same as or similar to that for collecting and preserving
reference and test product. For example, a volume of reference
product may comprise 200 liters packaged in 80 1-liter bottles with
Teflon.RTM.-lined caps, and in 480 250-ml bottles with Teflon-lined
caps. For laboratory-only testing, a volume of prototype product
may comprise 200 liters packaged in 800 250-ml bottles with
Teflon.RTM.-lined caps. For laboratory and on-line testing, a
volume of prototype product may comprise 3200 liters packaged in
800 250-ml bottles with Teflon.RTM.-lined caps, and comprise 2,400
liters in bulk storage in accordance with ASTM D2885 A2, and piped
to an on-line spectrographic analyzer.
[0116] In some embodiments, all products may be determined to be
homogeneous, where representative samples are shown to be
chemically the same throughout. If the reference fuel or prototype
fuel is taken from a blend header during a blend, the reference
fuel or prototype fuel may be taken into one large container (e.g.
a capture tank,) which then could be mixed and deemed homogeneous
and representative. Once the capture tank is mixed, lab samples may
be collected into chilled bottles through a chiller, the fuel
sample temperature brought to below 10.degree. C. (50.degree. F.,)
and a dip tube used that reaches the bottom of the bottles to avoid
agitation due to splashing. The bottles may be sealed with
Teflon.RTM.-lined caps to avoid loss of vapor pressure, labeled
with its representative grade, type and season and the date of
capture, stored out of direct light of any kind, and kept in an
environment where the temperature is no greater than 25.degree. C.
(77.degree. F.).
[0117] If the product is to be taken from a tank or a finished
product discharge pump, the recommended primary capture device may
be a clean (preferably new) stainless steel drum or drums, or a
portable semi-bulk container. Capturing the product may involve
pre-cooling the empty drum so that its skin temperature is less
than 10.degree. C. (50.degree. F.), filling the drum through a
chiller to cool the product to less than 10.degree. C. (50.degree.
F.), sealing the drum so as to leave approximately 2.5 cm of head
space, and storing the drums in an environment where the
temperature is no greater than 25.degree. C. (77.degree. F.). When
filling from the drum(s) to bottles or to a prototype fuel tank,
the bottles or tanks may be chilled and a drum pump may be used to
transfer fluid from the drum to the bottles or tanks. Samples may
be taken at the beginning, middle and end of the transfer process.
Those samples may be tested using a globally-calibrated
spectrometer to verify that the samples are homogenous.
[0118] To certify a prototype product against a reference product
for which consensus values have been determined, a direct match
comparison technique may be used in the same way as described above
for comparing a test product to a reference product.
[0119] A chemically-similar test product and reference product may
be tested together by spectrographic analysis. In the example of
fuel, to certify a chemically-similar prototype fuel against a
reference fuel, a direct match comparison technique may be used.
For each certification, a portion of the preserved reference fuel
sample may be obtained from a sealed container or from a piston
type or water displacement system. Both the reference fuel and
prototype fuel may be chilled in the same way prior to testing. For
example, the temperature may be greater than 0.degree. C.
(32.degree. F.) but not exceed 10.degree. C. (50.degree. F.). All
comparisons may be accomplished using a single globally-calibrated
spectrometer. Alternatively, different globally-calibrated
spectrometers may be used. The reference fuel and prototype fuel
may be tested in sequence for one or more cycles, e.g., 10
cycles.
[0120] The direct match comparison technique may further comprise
calculating the delta between the spectrographically-determined
values for the prototype fuel and the reference fuel with respect
to each property:
.DELTA..sub.i=PI.sub.i-RI.sub.i (Equation 3)
Where:
[0121] .DELTA..sub.i=difference between the
spectrographically-determined property values of the prototype fuel
and the reference fuel [0122]
PI.sub.i=spectrographically-determined property value of the
prototype fuel (text index) [0123]
RI.sub.i=spectrographically-determined property value of the
reference fuel (reference index) i=a property, such as MON or
RON
[0124] The method may further comprise calculating the derived
property of the prototype fuel (P.sub.i):
P.sub.i=.DELTA..sub.i+R.sub.i (Equation 4)
Where:
[0125] P.sub.i=derived property value of the prototype fuel for
certification [0126] R.sub.i=property value of the reference fuel
determined by primary testing (e.g., consensus value, laboratory
reference value or semi-consensus value)
[0127] Thus, the direct match comparison technique may comprise
adding the delta (.DELTA..sub.i) to the property value of the
reference fuel determined by primary testing (R.sub.i) to establish
the property value of the prototype fuel that may be used for
certification against the reference fuel. In various embodiments,
R.sub.i may comprise the consensus value, the laboratory reference
value or semi-consensus value of the reference fuel.
[0128] FIG. 5 illustrates an exemplary embodiment of a method 55
for direct match comparison of a prototype fuel against a
chemically-similar reference fuel. In a step 56, a sample of a
prototype fuel for which analysis of a property value is desired
may be collected. In a step 58, a reference fuel may be selected
for use in a direct match analysis of the prototype fuel. The
prototype fuel may be chemically similar to the reference fuel.
Chemical similarity may be determined in the manner discussed in
more detail above.
[0129] In a step 60, a refinery's laboratory may run a
spectrographic test of the prototype fuel and the reference fuel
using a globally-calibrated spectrometer. The selected reference
fuel and the prototype fuel may be measured in a spectrographic
analyzer any number of times. As shown in Table 4 (below),
sequential spectrographic measurements of the prototype and
reference sample may be executed 10 times.
[0130] In a step 62, the deltas between the determined property
values of the reference fuel and the prototype fuel may be
calculated and averaged. For example, the average delta for the RON
value associated with ten sequential measurements of the prototype
and reference sample in this example may be 0.14. The average delta
for the MON values may be 0.24.
[0131] In step 64, the average delta values may be added to the
consensus value of a property for the reference sample. For
example, in the example in Table 4, consensus values of 92.16 for
RON and 82.25 for MON may have been previously determined, such as
by measurement of the reference sample by a number of external
laboratories. Addition of the average deltas to the consensus value
may then be used to determine the certified value of the prototype
fuel. For example, for RON addition of the average delta value of
0.14 to the RON consensus value of 92.16 yields a certified value
of 92.30. For MON addition of the average delta value of 0.24 to
the MON consensus value of 82.25 yields a certified value of
82.49.
TABLE-US-00004 TABLE 4 Spectroscopic Spectroscopic RON RON
Spectroscopic Spectroscopic MON Prototype Fuel Reference Fuel RON
Delta MON PROTO Reference Fuel MON Delta 91.9 91.88 0.02 82.39
82.15 0.24 92.14 92 0.14 82.31 82.14 0.17 92.15 91.89 0.26 82.47
82.25 0.22 91.9 91.8 0.1 82.34 82.23 0.11 92.07 91.82 0.25 82.23
82.00 0.23 92.03 91.94 0.09 82.33 82.22 0.11 92.13 91.88 0.25 82.38
82.00 0.38 91.94 91.94 0 82.46 82.12 0.34 91.89 91.87 0.02 82.40
82.00 0.40 92.07 91.78 0.29 82.34 82.19 0.15 Avg. = 92.02 Avg. =
91.88 0.14 Avg. = 82.36 Avg. = 82.13 0.24 Standard 92.16 Standard
Consensus 82.25 Consensus Value Value Certified Value 92.30
Certified Value 82.49
[0132] Once a prototype fuel has been certified, it may be used for
testing against a chemically similar test product having unknown
property values. For example, as discussed above, a prototype fuel
may be used for on-line analysis. Thus, a test product may be
tested against either or both of a reference or prototype
product.
[0133] FIG. 6 illustrates an exemplary embodiment of a method 65
for direct match comparison of a test fuel against a
chemically-similar reference or prototype fuel. In a step 66, a
sample of a test fuel for which analysis of a property value is
desired may be collected. In a step 68, a reference or prototype
fuel may be selected for use in a direct match analysis of the test
fuel. The test fuel may be chemically similar to the reference or
prototype fuel. Chemical similarity may be determined in the manner
discussed in more detail above. In a step 70, a refinery's
laboratory may run a spectrographic test of the test fuel and the
reference or prototype fuel using a globally-calibrated
spectrometer. In step 72, the differences, or deltas, between the
spectrographically-determined test and reference or prototype
values may be calculated. In step 74, the property value deltas may
be added to the reference consensus property values to derive the
property values of the test fuel suitable for certification.
[0134] FIG. 7 illustrates one embodiment of a system 76 for direct
match comparison. The system may comprise a spectrographic analyzer
78 capable of wireless or wired communication with a computer 80
and server 82 via network 84. The server 82 may be connected to a
database 86. The spectrographic analyzer 78 may be an on-line
analyzer connected so as to allow collection of a sample 88 of a
product stream 90. Alternatively, the spectrographic analyzer 78
may be a standalone device in a refinery laboratory or an outside
laboratory. The spectrographic analyzer 78 may be
globally-calibrated as described above. In other embodiments, for
example, global calibration data may be stored in the database 86.
Reference values (e.g., consensus, semi-consensus and laboratory
reference values) determined by primary testing, such as engine
testing, may be stored in the database 86.
[0135] For direct match comparison of a test product to a reference
or prototype product, or comparison of a prototype product to a
reference product, the spectrographic analyzer 78 may be used to
spectrographically determine one or more property values of the
test, prototype and/or reference product, as the case may be. In
the example of a direct match comparison between a test product and
a reference product, the spectrographically-determined reference
property value and spectrographically-determined test property
value may be communicated to the computer 80, which may be
programmed with the algorithms discussed above to determine the
delta between the two values. The computer 80 may be further
programmed to request and receive a reference consensus value for
the property from the database 86 via server 82. The computer 80
may be further programmed to add the delta to the consensus value
to derive the certified value of the test product.
[0136] The computer 80 may also be programmed to determine a
laboratory reference value or a semi-consensus reference value, as
described above. In some embodiments, the spectrographic analyzer
78 (which may or may not be globally calibrated for this example)
may be used to perform multiple property value tests on a reference
product sample, and send the test data to the computer 80. The
computer 80 may be programmed to statistically treat the data as
described above to determine a laboratory reference value. In other
embodiments, the computer 80 may be programmed to combine the test
data from the spectrographic analyzer 78 with data from outside
laboratories, and statistically treat the data as described above
to determine a semi-consensus reference value.
[0137] In the embodiment of FIG. 7, the system 76 may comprise a
client-server architecture. Of course, the system 76 may comprise
more than one server and/or client. In other embodiments, the
system 76 may comprise other types of network architecture, such as
a peer-to-peer architecture, or any combination or hybrid
thereof.
[0138] The server 82 may comprise one or more computers or programs
that respond to commands or requests from one or more other
computers or programs, or clients. The client 80 may comprise one
or more computers or programs that issue commands or requests for
service provided by one or more other computers or programs, or
servers. The server 82 and client 80 may be located in or across
one or more computers and/or geographic locations. Servers and/or
clients may variously be or reside on, for example, mainframe
computers, desktop computers, PDAs, smartphones (such as Apple's
iPhone.TM., Motorola's Atrix.TM. 4G, and Research In Motion's
Blackberry.TM. devices), tablets, netbooks, portable computers,
portable media players with network communication capabilities
(such as Microsoft's Zune HD.TM. and Apple's iPod Touch.TM.
devices), cameras with network communication capabilities, wearable
computers, point of sale devices, spectrographic analyzers, and the
like.
[0139] A computer may be any device capable of accepting input,
processing the input according to a program, and producing output.
A computer may comprise, for example, a processor, memory and
network connection capability. Computers may be of a variety of
classes, such as supercomputers, mainframes, workstations,
microcomputers, and PDAs, according to the computer's size, speed,
cost and abilities. Computers may be stationary or portable, and
may be programmed for a variety of functions, such as cellular
telephony, media recordation and playback, data transfer, web
browsing, data processing, data query, process automation, video
conferencing, artificial intelligence, and much more.
[0140] A program may comprise any sequence of instructions, such as
an algorithm, whether in a form that can be executed by a computer
(object code), in a form that can be read by humans (source code),
or otherwise. A program may comprise or call one or more data
structures and variables. A program may be embodied in hardware or
software, or a combination thereof. A program may be created using
any suitable programming language, such as C, C++, Java, Perl, PHP,
Ruby, SQL, and others. Computer software may comprise one or more
programs and related data. Examples of computer software include
system software (such as operating system software, device drivers
and utilities), middleware (such as web servers, data access
software and enterprise messaging software), application software
(such as databases, video games and media players), firmware (such
as software installed on calculators, keyboards and mobile phones),
and programming tools (such as debuggers, compilers and text
editors).
[0141] Memory may comprise any computer-readable medium in which
information can be temporarily or permanently stored and retrieved.
Examples of memory include various types of RAM and ROM, such as
SRAM, DRAM, Z-RAM, flash, optical disks, magnetic tape, punch
cards, EEPROM. Memory may be virtualized, and may be provided in or
across one or more devices and/or geographic locations, such as
RAID technology.
[0142] An I/O device may comprise any hardware that can be used to
provide information to and/or receive information from a computer.
Exemplary I/O devices include disk drives, keyboards, video display
screens, mouse pointers, printers, card readers, scanners (such as
barcode, fingerprint, iris, QR code, and other types of scanners),
RFID devices, tape drives, touch screens, cameras, movement
sensors, network cards, storage devices, microphones, audio
speakers, styli and transducers, and associated interfaces and
drivers.
[0143] A network may comprise a cellular network, the Internet,
intranet, local area network (LAN), wide area network (WAN),
Metropolitan Area Network (MAN), other types of area networks,
cable television network, satellite network, telephone network,
public networks, private networks, wired or wireless networks,
virtual, switched, routed, fully connected, and any combination and
subnetwork thereof. The network may use a variety of network
devices, such as routers, bridges, switches, hubs, repeaters,
converters, receivers, proxies, firewalls, translators, and the
like. Network connections may be wired or wireless, and may use
multiplexers, network interface cards, modems, IDSN terminal
adapters, line drivers, and the like. The network may comprise any
suitable topology, such as point-to-point, bus, star, tree, mesh,
ring, and any combination or hybrid thereof.
[0144] Communication in and among computers, I/O devices and
network devices may be accomplished using a variety of protocols.
Protocols may include, for example, signaling, error detection and
correction, data formatting and address mapping. For example,
protocols may be provided according to the seven-layer Open Systems
Interconnection model (OSI model), or the TCP/IP model.
[0145] In other embodiments, a computer may be programmed to store
primary test reference values in its memory and/or retrieve such
values from a database. The computer may be programmed to receive
(1) a property value of a first refinery product, such as a
reference fuel, the property value determined using a
non-spectrographic test, (2) a property value of the first refinery
product determined using a spectrographic analyzer, and (3) a
property value of a second refinery product determined using the
spectrographic analyzer. The computer may be further programmed to
determine a difference between the spectrographically-determined
property values of the first refinery product and the second
refinery product; and to add the difference to the
non-spectrographically-determined property value of the first
refinery product to derive a property value for the second refinery
product.
[0146] In yet other embodiments, a spectrographic analyzer may be
configured to receive a property value of a first refinery product
using a non-spectrographic test; determine a property value of the
first refinery product using a first globally-calibrated
spectrographic analyzer; determine a property value of a second
refinery product; determine a difference between the
spectrographically-determined property values of the first refinery
product and the second refinery product; and add the difference to
the non-spectrographically-determined property value of the first
refinery product to derive a property value for the second refinery
product.
[0147] In yet other embodiments, a system may be provided for
direct match comparison. The system may comprise a spectrographic
analyzer capable of determining a property value of a first
refinery product, and determining a property value of a second
refinery product. The system may further comprise a computer
programmed to receive a property value of the first refinery
product, the property value determined using a non-spectrographic
test, and to receive the property values of the first and second
refinery products determined by the spectrographic analyzer. The
computer may be further programmed to determine a difference
between the spectrographically-determined property values of the
first refinery product and the second refinery product, and add the
difference to the non-spectrographically-determined property value
of the first refinery product to derive a property value for the
second refinery product.
[0148] Improved precision (e.g., reduced test variation) provided
by use of spectrometers, may result in a variety of benefits to
refineries. One benefit may be better risk management. Improved
precision may significantly lower the probability of shipping a
product that does not materially conform to product specifications,
even if tests of the product show material conformance with such
specifications. For example, improved precision may result in
reducing test variation from approximately 1 error in approximately
20 tests to approximately 1 error in approximately 1000 tests.
[0149] Another benefit may be better economics. Reduced variation
may allow the product producer to better target a product property
value more closely to a product specification, thus reducing
property giveaway.
[0150] When a spectrographic analysis is done, the resulting data
for a test fuel may be tested against a reference fuel that may
have similar properties (chemical similarity) to the fuel being
tested. In some embodiments, the test fuel may have certified
properties that are no more than one degree of separation from a
reference fuel, i.e., the test fuel is referenced against a
prototype fuel that may itself be tested against a reference fuel.
The test result of an unknown fuel may be the determined property
value of the test fuel, which may be calculated by subtracting by
the spectrographically-determined property of the reference fuel
from the spectrographically-determined property of the test fuel
(the A), and adding the delta to the consensus value of the
reference fuel.
[0151] In some cases, it may be determined that a particular
spectrometer technology is very stable in a laboratory environment.
In this case, over time, when repeated tests are run on the same
fuel (e.g. a laboratory control standard), essentially the same
spectrum is consistently produced. In this case, and where a
particular laboratory can produce accurate and precise reference
test results, that laboratory can create a library of reference
data and spectra. When a sample is then tested, the program looks
up the best fit spectrum, and uses that matched data for the
reference fuel function.
[0152] Although the present invention and its advantages have been
described in detail, it should be understood that various changes,
substitutions and alterations can be made herein without departing
from the invention as defined by the appended claims. Moreover, the
scope of the present application is not intended to be limited to
the particular embodiments of the process, machine, manufacture,
composition, or matter, means, methods and steps described in the
specification. As one will readily appreciate from the disclosure,
processes, machines, manufacture, compositions of matter, means,
methods, or steps, presently existing or later to be developed that
perform substantially the same function or achieve substantially
the same result as the corresponding embodiments described herein
may be utilized. Accordingly, the appended claims are intended to
include within their scope such processes, machines, manufacture,
compositions of matter, means, methods or steps.
* * * * *