CN1430663A - Crude oil blending method and system - Google Patents

Abstract

A method for blending two or more constituents into a petroleum mixture employs nuclear magnetic resonance to obtain real-time estimates of selected properties of at least one of the constituents. A multivariate controller processes these estimates to determine the relative amounts of each constituent that are required to form a petroleum mixture having desired values of those selected properties.

Description

Crude oil blending method and system

The present invention relates to a crude oil refining method and system, in particular to a method and system for blending various crude oils to obtain a blended petroleum mixture with desired physicochemical properties.

Background of the invention

Crude oil, also known as petroleum, is a complex mixture of hydrocarbons. These constituent hydrocarbons are separated from one another during the process of making commercially useful petroleum products. The various physical and chemical processing steps that separate crude oil into its component hydrocarbons are collectively referred to as "refining".

A difficulty associated with crude oil refining is the variability in the properties of the crude oil. For example, crude oil exists at room temperature with a similar consistency to heavy motor lubricating oil. Crude oil with bituminous consistency is also present at the same temperature. Although crude oil is generally considered to be black, crude oil may be brown, yellow, green, or red. There are even crude oils that fluoresce. These differences are a manifestation of the large number of different hydrocarbon mixtures found in different fields around the world.

The physicochemical processes involved in refining crude oil depend largely on the physicochemical properties of the oil. These properties in turn depend on the constituent hydrocarbons of the petroleum and their respective concentrations. Because the composition of crude oils is extremely different, the degree of variation in these refining processes makes it impossible to easily build refineries optimized for each of the different crudes. As a result, it is common practice in the refining industry to blend several crude oils to form a petroleum mixture suitable for processing in the refinery.

The process of blending several crude oils to produce a petroleum mixture suitable for a particular refinery requires up-to-date knowledge of the values of selected properties of the various crude oil components of the mixture. These selected properties include aromaticity, boiling point, flash point, cloud point, viscosity, pour point, API gravity, freezing point, octane number, PIONA, and RVP of the crude oil. However, gaining quantitative knowledge of these properties is not a simple matter.

The above properties of crude oil can be obtained by gas chromatographic analysis. However, this method is impractical for heavy crude oils because the heavier components present in such crude oils are not sufficiently volatile to be desorbed from the chromatographic column. These heavier components tend to remain in the column, thus making the column unusable for further testing.

Another way to obtain the physicochemical properties of crude oil is to observe the interaction of crude oil with infrared radiation. However, this method is sensitive to the opacity of crude oil. Therefore, this method is not effective for heavier, highly opaque crudes. Additionally, the interaction of crude oil with infrared radiation is highly nonlinear and temperature sensitive.

Another way to obtain the physical and chemical properties of crude oil is to carry out laboratory tests on samples. However, this method is costly and time-consuming. Because the types of crude oil used for blending vary from tank to tank, it is difficult to perform laboratory testing frequently enough to maintain current information at any time on the properties of all the types of crude oil used in the blend. Therefore, this method is not suitable for real-time control of the numerical values of selected properties of petroleum mixtures.

Accordingly, it is an object of the present invention to provide a method and system for obtaining up-to-date information on the values of selected properties of one or more types of crude oils and to provide methods for blending those crudes in order to obtain petroleum mixtures having desired values for those selected properties. methods and systems.

Summary of the invention

The method of the present invention overcomes the deficiencies of the prior art by performing real-time, on-line evaluation of the value of at least one selected property of the constituent crude oils used to blend the petroleum blend. The results of this real-time assessment are provided to the automatic controller along with the desired petroleum blend properties. In real-time monitoring, the controller calculates the appropriate amount of constituent crude oil needed to produce a petroleum blend having desired values of selected properties. Since the evaluation is performed in real time, the system incorporated into the present invention dynamically changes the relative amounts of the components of the blended petroleum mixture as the values of selected properties of each constituent crude oil change. The term "real time" is used herein in a relative sense, with time generally measured in seconds or minutes. Such timing satisfies crude blending controls on a commercial basis.

The preferred practice of the invention uses nuclear magnetic resonance measurements associated with numerical assessments of selected properties of each constituent crude oil. NMR measurements do not rely on optical or infrared radiation and are therefore not affected by high opacity. In addition, NMR measurements are relatively stable with temperature. As a result, reliable measurements can be made at the higher temperatures to which crude oil is typically heated in order to pass it through the pipeline.

One embodiment of the invention includes the step of applying a steady magnetic field to a sample of the first constituent crude oil. With a steady magnetic field in place, the NMR sensor applies a transient magnetic field to a crude oil sample of the first constituent and measures the response of the first constituent to the transient magnetic field. NMR sensors generally do not directly provide values for selected properties of the first constituent crude oil. However, NMR sensors provide information about the chemical composition of the first constituent. More specifically, NMR measurements provide the chemical spectral signature of hydrogen present in a sample of the material being tested. To this end, the method of the invention comprises the step of evaluating the properties of the first constituent on the basis of its NMR measurements. Accordingly, the computer system estimates a value for a selected property of the first constituent based on such a response. These properties are then used to selectively blend the first constituent with the second constituent to form a blended petroleum mixture having the desired values of the selected properties. Examples of such selected properties include aroma, boiling point, flash point, cloud point, viscosity, pour point, API gravity, freezing point, octane, PIONA, and RVP.

Expected values for selected properties of the petroleum mixture are generally specified by an optimization program based on constraints imposed by the particular refinery that processes the blended petroleum mixture. However, the choice of these expectations may also be influenced by economic factors such as the prices and availability of different crude oils and the selling prices and demand for various products refined from blended petroleum mixtures.

The method of the present invention may be implemented with a feedback system that evaluates the values of selected properties of the blended petroleum mixture and compares those estimates to expected values of those selected properties. The differences between the estimated values of the selected properties and the expected values of those properties are then used to adjust the relative amounts of the various constituents of the blended petroleum mixture. Preferably, NMR-response information relating to the composition and more particularly the hydrogen chemistry of the blended mixture is first determined using nuclear magnetic resonance, and then the numerical evaluation of the selected property of the blended petroleum mixture is performed from the measured information to evaluate the value of the selected property of the mixture.

The method of the present invention can thus provide real-time assessments of selected property values for various crude oils. Since the method of the present invention relies on NMR rather than optical techniques, the precision of these assessments is essentially independent of the opacity or temperature of the crude oil. As a result, the method of the present invention is eminently suitable for blending various crude oils into blended petroleum mixtures having selected desired property values.

A system practiced with the present invention includes an optimization program for specifying desired values for selected properties of the petroleum mixture formed. The optimization procedure specifies expectations based on refinery characteristics, optionally based on economic factors.

The system also includes a sensor for assessing the value of at least one selected property of the constituents for blending into the mixture. These estimates are provided to the controller along with the expectations specified by the optimizer. Based on the evaluations from the sensors and the desired values from the optimization routine, the controller determines the relative amounts of constituents required to form a petroleum mixture with desired values for selected properties.

Using the above techniques as further described below, the method and apparatus according to the present invention are capable of minimizing the variability of the crude oil mixture delivered to the refinery. Additionally, the mixture is optimal for a particular refinery.

These and other features and advantages of the present invention are apparent from the following detailed description and accompanying drawings, in which,

Brief description of the drawings

Figure 1 shows a feed-forward crude oil blending control system embodying features of the present invention;

Figure 2 shows details of the control system of Figure 1; and

Figure 3 shows a modification of the control system of Figure 2, configured to operate as a feedback control system.

Detailed description of the invention

A petroleum blending system 10 incorporating the principles of the present invention receives feed from an oil depot 12 having a plurality of oil storage tanks 14a-n, three of which are shown in FIG. 1 . These feeds are typically different classes of crude oils, each characterized by an input vector of properties. These input vectors are denoted in Figure 1 by symbols x ₁ , x ₂ , . . . x _n associated with the respective tanks. The elements of a typical input vector _xi include values representing selected physicochemical properties of the crude oil. The output of the blending system 10 is a blended mixture of various crude oils from the depot 12 . This blend mixture of crude oils is characterized by an output vector y having elements representing the chemical physical properties of the blend mixture. Examples of such selected properties include viscosity, aroma, API gravity, etc.

While Figure 1 illustrates various crude oils that are transported by pipeline from depots to refineries, other sources of crude oil are available. For example, one or more types of crude oil can travel directly from a wellhead or from a storage tank to a refinery via a pipeline. Alternatively, multiple crude oils may arrive at the well from different depots.

The petroleum blending system 10 selects the quantities of various crude oils required to obtain a blended mixture with an optimized output vector, eg, to maximize refinery revenue. The output of the optimization is given by the setpoint vector r provided by the optimization program 13 . This setpoint vector r includes elements representing expected values for selected properties of the blended petroleum mixture. The optimization program 13 determines a setpoint vector according to the constraints imposed by the properties of the particular refinery that will obtain this blended petroleum mixture. In addition, the optimizer 13 can use current and expected economic conditions to determine the setpoint vector. Such economic conditions may include the purchase price and availability of various types of crude oil, the selling price and demand for various products made from the crude oil, and the costs associated with the manufacture of these products.

The optimizer 13 is typically implemented as software instructions executing on a programmable digital processor, such as a general purpose digital computer. The information is provided to the optimizer 13 by an operator using a keyboard. Alternatively, optimizer 13 may receive information via a network connection. In one practice of the invention, the optimization program 13 is set to monitor a global computer network, such as the Internet, for the key economic indicators used in determining the setpoint vector r. A suitable optimization program for the practice of the present invention is sold under the trade name ROMEO by Simulation Sciences of Brea, California.

Petroleum blending system 10 is thus a multivariable control system that operates based on a number of input vectors x ₁ , _{x 2} , . The output vector y includes elements representing actual values of selected properties of the petroleum mixture. The control system shown in Figure 1 is a feed-forward system because the output vector y is not fed back as an input. However, as discussed above, the invention is equally applicable to feedback control systems.

FIG. 2 shows details of the petroleum blending system 10 of FIG. 1 . As shown in FIG. 2 , the first oil storage tank 14 a is connected to the first pipeline 16 for transporting the first crude oil from the oil depot 12 to the blending station 18 . The first pipeline 16 includes a first oil pump 20 to propel the crude oil through the pipeline 16 and a first valve 22 to control the amount of crude oil delivered to the blending station 18 . The position of the first valve 22 is controlled by a first regulator 23 .

The second oil storage tank 14b from the oil depot 12 is also connected to the second pipeline 26 for transporting the second crude oil to the blending station 18 . The second pipeline 26 includes a second oil pump 28 to propel the crude oil through the pipeline 26 and a second valve 30 to control the amount of crude oil delivered to the blending station 18 . The position of the second valve 30 is controlled by a second regulator 31 .

Between the first storage tank 14 and the blending station 18 , a first nuclear magnetic resonance (NMR) sensor 32 is fitted to sample the crude oil in the first pipeline 16 through a first shunt 33 . Likewise, between the second oil storage tank 24 and the blending station 18 , a second NMR sensor 34 is installed to collect a sample of the crude oil in the second pipeline 26 through the second shunt pipe 35 .

A preferred NMR sensor 32, 34 uses technology from the I/A Series NMR apparatus of Foxboro Company, Massachusetts, USA.

The outputs of the first and second NMR sensors 32 , 34 are provided to first and second stoichiometric modeling units 40 , 42 . These stoichiometric simulation units 40, 42 convert the output of the NMR sensors 32, 34 into a format suitable for the multivariable controller 36 associated with the simulation units. In response, the multivariable controller 36 sends control signals to the first and second regulators 23 , 31 . The first and second regulators 23, 31 control the position of the valves 23, 30 in such a way that the output vector y is within the tolerance range of the setpoint vector r provided by the optimization program 13, the output vector y representing the regulation properties of the mixture. These properties may include aroma, boiling point, flash point, cloud point, viscosity, pour point, API gravity, freezing point, octane number, PIONA, and RVP.

A suitable stoichiometric modeling unit can be implemented on a programmable digital processor as a lookup table or as a mathematical model. A stoichiometric simulation cell can be placed close to the NMR sensor, as shown in Figure 2. Another optional practice is for the stoichiometric modeling unit to be in the multivariate controller 36 . However, the NMR sensors 10 in the petroleum blending system 10 may also share a common stoichiometric modeling unit.

The optimizer 13 and multivariable controller 36 and stoichiometric modeling unit are preferably implemented as software executing on a programmable digital processor. In practice, these instructions are executed on a general purpose digital computer. To meet desired performance requirements, the optimization program 13 and the multivariable controller 36 can be implemented in hardware, software, or a combination of hardware and software. A multivariable operating controller suitable for the practice of the present invention is sold under the tradename CONNOSIEUR by Simulation Science Inc. of Brea, California. The detailed implementation details of the multivariable controller 36 and the optimization program 13 are known to those skilled in the art and do not affect the scope of the present invention.

While the petroleum blending system 10 illustrated in FIG. 2 shows two distinct NMR sensors 32, 34, it will be appreciated that a single NMR sensor may be used for the first and second storage tanks on a time-sharing basis. However, depending on the details of the refinery layout, the distance between the various storage tanks may make it impractical to interface a single NMR sensor with two or more storage tanks. In practice, a single NMR sensor can serve as one, two, or multiple sources of crude oil, depending on the refinery layout and constraints imposed by the infrastructure.

In operation of the petroleum blending system 10 illustrated in FIGS. 1 and 2 , the first oil pump 20 propels petroleum through the first conduit 16 . A sample of the petroleum is diverted to a first NMR sensor 32 through a first split tube. The rest of the oil flows to the first valve 22 and, to the extent that the first valve is open, reaches the blending station 18 .

NMR sensor 32 applies a steady magnetic field across the sample to orient the magnetic dipole moment associated with the molecules in the sample. When the stable magnetic field is in place, the NMR sensor 32 then applies a transient magnetic field in a different direction than the stable magnetic field, preferably a transient magnetic field orthogonal to the direction of the stable magnetic field. This transient magnetic field temporarily aligns the dipoles in the sample in a direction other than those in which the static magnetic field orients the dipoles. When the transient magnetic field is turned off, the dipoles in the sample jump back to the orientation imposed on them by the steady magnetic field. In this way, the dipoles generate an RF signal. The proportion of jumping back to a stable magnetic field orientation and thus the frequency of the resulting RF signal is characteristic of the molecular species carrying this dipole. The RF spectra thus generated provide a means of determining information about the chemical composition of the sample, more particularly the hydrogen chemistry of the sample.

Thus, NMR sensor 32 provides information about the chemical composition of the crude oil sample in pipeline 16 . Estimating corresponding values of selected properties from such measured sample composition response information is known in the art. This conversion of measured sample information into selected property estimates is performed by a first stoichiometric modeling unit 40 connected to the multivariable controller 36 and the NMR sensor 32 . Information input to the first stoichiometric simulation unit 40 is sample response information measured by the NMR sensor 32 . The output of the first stoichiometric unit 40 is a corresponding set of selected property assessments. The second NMR sensor 34 and the stoichiometric simulation unit 42 operate in the same manner as the first NMR sensor 32 and the stoichiometric simulation unit 40 .

A stoichiometric modeling unit 40 suitable for the practice of the present invention is implemented by a digital processor executing instructions to evaluate selected physical property values based on measured hydrogen chemistry of a crude oil sample. These instructions perform procedures known to those of ordinary skill in the art. These procedures include building a checklist, interpolating between two values in a checklist, or executing a mathematical model that evaluates selected property values.

Although not required for the practice of the invention, it is generally desirable to place a third NMR sensor 44 at the outlet of the blending station 18, as shown in FIG. 3, along with an associated third stoichiometric modeling unit 46. The third NMR sensor 44 and its associated third stoichiometric modeling unit 46 operate in the same manner as the combined first NMR sensor 32 and its associated stoichiometric modeling unit 40 discussed above. The output of the third stoichiometric modeling unit 46 is fed back to the multivariable controller 36, said output representing an estimate of selected property values of the blended petroleum mixture. When the third NMR sensor 44 is in place, for example, the multivariable controller 36 may detect anomalies in the blended petroleum mixture, which may indicate a system malfunction. In addition, the output of the third NMR sensor 44 may provide feedback variables that the multivariable controller 36 processes to augment these variables to produce control variables, such as valve positions, to achieve the desired oil mixture.

As shown in Figures 2 and 3, each of the NMR sensors 32, 34, 44 has an independent stoichiometric simulation unit 40, 42, 46 associated therewith. However, the blending system 10 can also be implemented with a single stoichiometric modeling unit coupled to each NMR sensor and used by each NMR sensor on a time-shared basis.

Although the invention is disclosed herein as applied to the blending of two components, it is evident from the context that the principles of the invention are readily extended to the blending of more than two components. It is also apparent that the blending system can be used to blend crude oil with materials of known composition.

Having described the invention and its preferred embodiments, what is claimed new and protected by Letters Patent is:

Claims (17)

1. A method of blending at least a first and a second component into a petroleum crude oil mixture, said method comprising the steps of: specifying desired properties of said petroleum-crude mixture, performing on-line NMR measurements of said first component to determine the measured properties of said first component, and Based on the measured properties, the first component is selectively blended with the second component to form a petroleum crude oil mixture having the desired properties. 2. The method of claim 1, wherein said step of performing on-line NMR measurements comprises: applying a steady magnetic field on said first component, superimposing a transient magnetic field on said steady magnetic field, measuring the response of said first component to said transient magnetic field, and Based on the response, a measured property of the first component is determined. 3. The method of claim 1, wherein said step of performing on-line NMR measurements comprises: determining information related to the composition of said first component, and Based on the composition information, the property of the first component is determined. 4. The method of claim 1, wherein the property comprises the chemical composition of the first component. 5. The method of claim 1, wherein said properties comprise physical properties of said first component. 6. The method of claim 5, wherein the physical property is selected from the group consisting of aroma, boiling point, flash point, cloud point, viscosity, pour point, API gravity, freezing point, octane number, PIONA, and RVP. 7. The method of claim 1, further comprising the step of specifying said expected value of said selective property based on market conditions. 8. The method of claim 1, wherein said step of selectively blending said first component and said second component comprises the steps of: performing on-line NMR measurements of said petroleum mixture to determine the measured properties of said petroleum mixture, determining the difference between said selected measured property of said petroleum mixture and said desired property, and Based on the difference, the amount of the first component included in the petroleum mixture is adjusted to reduce the difference. 9. The method of claim 8, wherein said step of performing an on-line NMR measurement of said petroleum-crude mixture comprises the steps of: applying a steady magnetic field to the petroleum mixture, superimposing a transient magnetic field on said steady magnetic field, measuring the response of said petroleum mixture to said transient magnetic field, and From the response, a measured property of the petroleum mixture is determined. 10. The method of claim 9, wherein said step of determining a measured property of said petroleum mixture comprises the steps of: determining the mixture composition of said petroleum mixture, and Based on the composition of the mixture, a measured property of the petroleum mixture is determined. 11. A system for blending at least first and second components into a petroleum mixture, said system comprising: an optimization procedure for specifying expected values for selected properties of said petroleum mixture, a first NMR sensor for on-line measurement of a property of said first component, and a controller coupled to the optimization program and the first NMR sensor, the controller selectively blending the first component with the second component based on the measured properties to form a oil mixture. 12. The system of claim 10, wherein the first NMR sensor comprises: means for measuring information responsive to said first component composition, and means for determining properties of said first component based on said composition information. 13. The system of claim 12, wherein said determining means includes means for determining information responsive to the chemical composition of said first component. 14. The system of claim 12, wherein said means for determining comprises means for determining a physical property of said first component. 15. The system of claim 14, wherein said means for determining a physical property of said component comprises a means for determining a component selected from the group consisting of aroma, boiling point, flash point, cloud point, viscosity, pour point, API gravity, freezing point, octane Values, means of physical properties of PIONA and RVP. 16. The system of claim 11, wherein said optimization program includes means for specifying said desired properties based on market conditions. 17. The system of claim 11, further comprising: a second NMR sensor for on-line measurement of properties of said blended petroleum mixture, and Feedback means for providing a link between said second NMR sensor and said controller.

Images