Crude oil was discovered in the early 1800s. The refineries built in the late 19th and early 20th centuries were designed mostly to process fewer types of crude from fewer sources than today. The crude oil supply became abundant as oil exploration activities started discovering huge oil reserves in the world. Crude quality was consistent with one crude oil source. As the world started using more oil following advances in the transport industry, refineries started facing problems in the constant supply and quality of crude oil from one source. Crude blending became a viable option to ensure that crude units received a consistent quality of crude.
The blog will discuss crude blending to facilitate consistent crude quality to the crude distillation units in the downstream refining industry. The supply and quality of crude from a single source is becoming scarce in today’s market and crude blending alleviates some of the problems.
Characteristics of crude oil
Crude oil is characterized by several properties that depend on the source of origin. Hence, crude oil prices are based on their characteristic properties. These qualities also affect the processing units and types of products produced in refining. Lighter crude will produce lighter, more gaseous products like butane, pentane, LPG, etc. Heavy crude will result in more viscous and heavy products like asphalt, wax, coke, etc.
Density – This property is described in terms of API (American Petroleum Institute) gravity and is related to the specific gravity of oil as follows: API Gravity = (141.5/Specific Gravity) – 131.5
API has units called API degrees. Specific gravity value is standardized at 60-degree Fahrenheit for the calculation of API. We can see that the lighter the oil in density, the greater its API degree and vice versa. Based on this criterion, crude oil is classified as follows:
- Light – API > 31.1
- Medium – API between 22.3 and 31.1
- Heavy – API < 22.3
- Extra Heavy – API < 10.0
Sulfur – Sulfur content is unavoidable and problematic in final fuel products such as gasoline, diesel, fuel, and bunker oil due to environmental regulations. Crude is termed as “sweet” for sulfur less than 0.7 wt% and as “sour” for sulfur more than 0.7 wt%. Sulfur is removed from crude oil through process units such as hydrodesulfurization (HDS) or hydrotreater and hydrocracking.
Total Acid Number (TAN) – This refers to the acidity of crude oil. An excessive value causes corrosion in process units. TAN is reduced in crude by the aqueous caustic base in presence of a catalyst. The TAN range in crude oil varies from 0.1 to 10 mg Potassium Hydroxide (KOH)/g oil. Viscosity – This property is important for the transportation of crude oil via pipelines to terminals or refineries.
Micro Carbon Residue (MCR) – This property measures the coke-farming tendencies of crude oil as it affects deposits in the combustion chambers of automobiles but is managed through the use of automobile additives. The range of MCR is 0-2 wt%.
Figure 1 shows the API and sulfur contents of 160 crude oil and their distinction as sweet/sour and light/heavy oil . We can see from Figure 1 that crude oil properties vary depending on their source of supply and may cause problems for the refining industry as refineries are designed to process stable qualities of crude. Crude blending is the best option to ensure a consistent quality of crude for the refinery. Since two crude properties, namely API and sulfur, are mostly considered for crude blending, it requires only two crude types to determine their ratios to meet the blended crude spec.
Advantages of crude blending
- Minimization of dependency on crude feed quality for stable and optimum operation
- Maximization of product yields and quality
- Improvement of crude oil inventory for “just-in-time” manufacturing
- Reduction in demurrage by better planning
- Traders use crude blending to meet pipeline specs for low-sulfur light crude to realize additional profits from the selling of crude to refiners.
Modes of crude blending
Crude blending is executed in many modes depending upon the source, destination, and sequential blending steps. Agrawal l  has discussed several of them. The most common mode is blending from pipelines to crude tanks in a refinery to feed the distillation units. Figure 2 shows a crude blending system that inputs 7 types of crude from pipelines to 5 paired intermediate crude storage tanks. A simple linear optimizer is used to optimize the crude ratios for each intermediate tank. This is step 1 of crude blending. Step 2 gives crude blend ratios from the intermediate crude tanks as feeds to different crude distillation columns.
Economics of crude blending
We have taken data for various crude types to evaluate the economics of crude blending.  We blended a 100KB crude tank from four types of crude, namely, light-sweet, medium-sweet, medium-sour, and heavy-sour to optimize crude ratios. We used a linear optimizer and objective function to minimize API and S specs of 37.5 (minimum) and 0.4 wt% (maximum) from the final blended crude qualities, respectively. Figure 3 shows the results of Excel-based optimization with a blended crude cost of $40.25/Bl of crude. This simple optimizer can optimize any number of crude qualities. Traders can use this methodology to estimate their profit based on cost and the selling price whereas refiners can use it to negotiate the price of blended crude. Ratios for two crude blending systems can be simply solved by iteration of nonlinear algebraic equations for API and sulfur.
Integration of product and crude blending
Agrawal  has discussed the modular components of the product and crude blending systems, which are shown in Figure 4. The benefits of crude blending cannot be realized unless it is integrated with other sub-systems for optimization, operational stability, and just-in-time manufacturing. These sub-systems can be categorized into two areas, namely, modules for crude quality control integration and modules for product quality control integration.
The above sub-systems are integrated only for the quality control of feed to the crude distillation unit and do not account for the optimum operation of the crude distillation unit and product blending optimization. Implementation of only crude quality control modules would require the targeted properties of the feed-to-crude unit as manual input to module-III shown in Figure 4. Thus, they would achieve only part of the benefits of the overall integration of both crude quality control and product quality control modules.
For example, target crude feed composition would depend upon the product blend optimization and target qualities of the side-streams of the crude unit. Feedback from the product quality control modules not only would compute the targeted crude feed composition but also set the targets for optimum operation of the crude unit.
Illustrative commercial crude blender
Crude blending at various points in the supply chain helps to enhance profits by reducing giveaways. Pipeline operators will also blend for viscosity reduction, thus reducing the horsepower required and providing energy cost savings and higher throughput.
For generations, this blending has occurred in a tank, but with advances in online analysis systems as well as control systems, the case for an inline blender is stronger because it has several important advantages over tank blending. For instance:
- Tanks are expensive to build, and motorized agitators have both operational and maintenance costs associated with them.
- Blending time can be substantial. Adding the time required for manual sampling, transport to the laboratory, and the laboratory analysis itself, the process can take days if not weeks.
- Obtaining a representative sample in a large storage tank is impossible.
- Correcting an off-spec blend requires a second blending period, manual sampling, transport, and lab analysis.
- Blending in the right product can create asphaltenes that will precipitate in the tank over time, requiring the expense of cleaning as well as the expense associated with an out-of-service tank.
The inline blender solves many of these problems:
- Blenders have a lower capital cost.
- The additional energy required to blend inline is minimal compared to motorized agitators due to low differential pressure.
- A wide variety of online analyzers measure the desired characteristics of the blend in or near real-time. Any off-spec result is corrected immediately.
- A properly designed mixing and analysis loop ensure a representative sample.
- When operated as a direct feed to the CDU, the feedstocks can be adjusted in minutes to optimize the refinery output.
In Figure 5, two or more streams are controlled for flow with valves and flowmeters to achieve a ratio target. The feedstocks are homogenized and then measured for one or more desired characteristics. The ratio target is adjusted accordingly to match the desired characteristics more closely. But as shown in Figure 6, the reality is a little more complicated.
One of the technologies that justifies an inline blender’s cost is the ultrasonic flowmeter. While not a new technology, recent improvements in accuracy, robustness, and cost make it ideal due to the availability in large sizes and the near-zero pressure drop making it suitable for gravity feed. Most other meters have a pressure drop associated with them that can make them unfeasible for retrofitting.
While each stream’s flow control valves are throttled in response to the measured flow, there is overall no flow control on an inline blender. For terminals and pipelines, the flow rate is the maximum allowed by the downstream backpressure. In a refinery, it is the flow control on the CDU itself that determines the flow rate. The inline blender therefore must adjust only the crude ratio in response to changes in downstream flow that it does not control.
Many inline blenders can use a wild stream, whereby one stream can run uncontrolled and the others are throttled to maintain the ratio. With crude blending, this is never feasible because this method requires that the pump pressure on the slave streams be greater than that on the wild stream. Therefore, the control system must balance each stream and does so by throttling each stream up and down respective to each other, paying close attention to not starve the downstream process.
For many blending operations, homogenization can occur during transit through the pipeline using its natural turbulence over some length to mix the feedstocks. However, our modern inline blender utilizes a variety of analyzers to control the desired properties of the crude. For these analyzers to be provided a representative sample, homogenization must be guaranteed to occur and do so in a short length.
Static mixers are employed in smaller systems. These devices are low-cost but poor fit for crude applications. For one, these devices have a pressure drop associated with them that increases their operating cost. They are also designed to operate in a narrow flow window and thus are ineffective at low flow ranges.
The proper method of homogenization in a crude blender is the jet mixer. This process involves a pump moving fluid downstream and injecting it a few meters upstream through a nozzle. The nozzle is designed to create enough turbulence to fully homogenize the contents of the main pipeline before it reaches the suction point. This side-stream loop is also a convenient location for the analyzers and other instruments.
The flow rate of the jet mix loop and the design of the injection nozzle are not something to be trivialized. Therefore, any inline blender should be using Computational Fluid Dynamics (CFD) to prove homogenization as shown in Figure 7.
Sampling and analysis
The above-mentioned homogenization becomes critical for the analysis (feedback) portion of the inline blender. Table 1 shows crude characteristics and their methods of analysis for a two-crude blending system.
Once the mechanics are properly designed, and the proper instrumentation is selected, the control system completes the inline blender. The control must first do no harm to the downstream process. Shutting off flow to a CDU will cause big problems.
A properly designed system will control the ratio of the blend without affecting the overall flow of the system. It will do this by PID algorithms coupled with gains to allow for differences in feed pressure and limits on valve closure. In our process, speed is not of the essence, and in fact, is a disruptor. Changes to the stream ratio must be performed without rushing to eliminate any potential to disrupt the downstream process.
An inline blend control system should include an optimizer that can consider each parameter to minimize giveaways and increase efficiency. Given the feedstocks, it may not be possible to fully optimize all the parameters. But a control system can be deployed that comes close.
As an example, the control system shown in Figure 8 displays several measured characteristics. Each of these is provided both an ideal target value and an acceptable range. The characteristics are provided a priority with #1 being the first-tier adjustment of the ratio. In our case, we set the actual ratio itself as priority #1 and give it a wide range. This ensures that in the event of a malfunction, the system will not command the ratio to some unreasonable position. Once adjusted to the target, the controller then looks at the parameter with priority #2 and adjusts the ratio to best achieve the 2nd target. If during the ratio change the higher priority characteristic falls out of range, the optimization is aborted. If both priority #1 and #2 can be optimized, the system then adjusts around priority #3 and so on.
For multi-stream blenders, the operator selects one stream as the control for each parameter. For example, Stream #1 is used to adjust the sulfur content. That channel’s ratio is then increased or decreased, and the other channels balance.
This blog discussed the characteristics of crude oil that are important for a crude blending system. Crude blending economics was illustrated by blending four light to heavy crudes and low to high sulfur contents. The tangible profit either for transfer or refinery was optimized using a nonlinear regression method.
Also, note that feedstocks can be incompatible with a major culprit being distillates into heavy crude. These interactions can cause the creation of asphaltenes that can be a big problem downstream. This incompatibility issue has been vastly studied and should be included in your crude blending project analysis.
- “Scope and Feasibility of Integrated Crude Blending Control and Optimization System”, S.S. Agrawal, ISA’94 Conference and Exhibit, Philadelphia, May 9-12, 1994
- “The Economics of Process Analysis”, G. Shahnovsky, A. Kiegel, R. McMurry, Hydrocarbon Engineering, Feb 2017
Dr. Suresh S Agrawal, CEO, Offsite Management Systems LLC
Tom Edwards, President, Technics, Inc.