In the oil and gas industry, there are generally two ways that we take measurements of product. The first is by volume. As fluid flows through a pipeline or sits in a tank, the industry has developed many ways by which we can capture that fluid and calculate “how much” of a particular product we have. These methods are calculated and refined to the point that we can determine these volumes down to the lowest possible unit. The second method by which we measure fluid is a quality measurement. Quality measurement has been well refined over the years to include new technologies not seen in the past. Electronic devices are very commonly used to provide real time analyzation of product as it flows through a custody transfer system. Although there have been many advances in the electronic quality measurement realm, another of the most common and widely accepted method of quality measurement is the composite sampling system. These systems incorporate the use of mechanical devices that capture small bites of the flowing product to compile a composite sample that is representative of the flowing batch. In crude oil sampling, these devices are used almost religiously to provide a primary source of measurement, or even a redundancy to an electronic system. In either scenario, there are certain considerations one should entertain in order to ensure that the composite sampling system is set up for success.
Composite sampling has been performed for years in our industry and has made much advancement over the course of the past thirty years. The driving force behind these advancements has been through the formation of standards for crude oil sampling and the individuals who sit on the standards boards and their contribution to the industry. There are three very well-known standards that are used worldwide for automatic sampling of crude oil. Two of these standards are identical in their recommendations and written by many of the same individuals sitting on both committees – those two standards are API 8.2 and ASTM D 4177. The third standard is ISO 3171. There are other standards as well, such as EI Section 2 Part VI, but the three aforementioned standards are the most frequently referred to in our industry. At the time of this paper, API section 8.2 has just been revised in late 2016 to include new practices deemed by the industry to be acceptable and beneficial to the art of sampling crude oil products. This document has also been expanded to include some recommendations on the sampling of other liquid petroleum products such as refined products. For the purpose of this paper, we will explore the recommendation and considerations that should be taken when developing an automatic sampling system for crude oil sampling.
When looking into building a crude sampling system that is sufficient for custody transfer, one must first consider the basic components and be sure that these are incorporated into the package. The first and most important component of the sampling system is the one that will determine all of the downstream components efficiencies before any of those components have a chance to operate. This of course is the means by which the sample stream is conditioned. Careful consideration must be placed on the type of conditioning or mixing that should be used prior to the collection of the sample. The three basic types of mixing systems are defined as; static mixing which can be achieved by placing angled elements inside of the pipeline to agitate the flowing stream, power mixing, which uses a combination of pump, motor and spraying nozzles inserted into the line to create a mixed stream and finally the use of piping bends and elbows. The use of piping bend and elbows has been deemed effective in the past, but is frequently less effective than the two previous mentioned methods. There are clear benefits to using any of the three, such as cost, effectiveness of mixing and application driven situations, the most widely accepted methods for mixing product prior to sampling is the static mixer or power mixer. One method that is met with some controversy as to whether it is very effective is the use of piping bends as a single source of pipeline mixing. Now, when using something such as an API loop for a custody transfer system, the manufacturer will often include a static mixing unit on the downstream side of the loop to ensure proper mixing. When considering the type of mixing apparatus to use for the sampling system, it is important to look at the application and consider the product and condition in which the mixing unit will operate. Density, viscosity, flow rate and gravity all will play a factor in the mixer you use. For example, the use of a static mixer in low velocity situation and a low API gravity crude oil may not yield the desired C1/C2 ratio and therefore not achieve a proper mix. Another example would be the use of a power mixer in a high velocity situation where the actual velocity of the product and piping configuration may produce an adequate mixture, rendering the power mixer unnecessary. Consider also the energy dissipation that may occur when using a static mixer that has too many elements causing too significant a pressure drop across the unit which may restrict the flow of product. In any scenario, the underlying objective of mixing the stream is to ensure that all the components in the sample stream are evenly mixed so that an accurate sample can be taken. This means that the mixing system, no matter which method used, must ensure that the water droplets are small enough to be sampled and the sediment in the flowing stream are uniformly distributed across the pipeline. There are obvious advantages and disadvantages to using the various styles of mixing systems available for crude oil sampling. While the answer to which conditioning method to use may not be clear, what is clear is that in order for the sampling system to yield results with any level of accuracy, the first step of a properly mixed stream must be achieved.
The second of these items to be considered, and possibly the most highly engineered of the sample system components, is the method by which the sample is collected. Of the few possible accepted methods to collect the sample, the most common of these is referred to as the sample extractor. The purpose of the extractor is to obtain a sample in a manner which is considered isokinetic. Isokinetic is defined as the ability to take a sample from the flowing stream where the condition is representative of the contents of the stream and is in normal flowing conditions. This is the best way to ensure the extractor will collect the contents of the stream and yield a high level of accuracy. The extractor itself consists of a chamber in which a known volume can be received, captured and then displaced to a container in a manner that is repeatable. This repeatability has a small allowable variance, but the best systems are capable of repeatability to a single digit percentage; usually less than 5% variance. The collection chamber is most commonly referred to as the collection head, or “grabber” portion of the sampler. This “grabber” can come is varying designs, any of which provides the user with a distinct advantage when collecting their sample. Sample extractors may also come in many different sizes and designs. Some of these extractors are equipped with insertion mechanisms that allow them to be inserted into pipelines up to 48” in diameter. Not only can these probes reach into the center of these large pipeline (center ½ of pipeline reach is required by API standards), but in some cases they can also be inserted into these pipelines under the extreme pressure of flowing conditions. These manual and auto insertion mechanisms can assist the user in maintenance situations where shutting down a pipeline due to inoperable equipment may not be suitable for contract standards. Another type of sample extractor, which is very similar to the main line sampler defined above, is called an in-loop or cell sampler. The in-loop sampler operates in a very similar fashion by grabbing the sample from a flowing stream. The inherent difference is that this particular sampler is not installed into a main line, but is installed into a small (usually 1” or 2”) bypass line that is taken off of the main pipeline. These in-loop samplers have the capability to produce similarly accurate results to that of the mainline sampler with proper set up and use. Because these units are actuated mechanical devices, they require a power source. Most commonly the grabber is powered by a pneumatic/hydraulic source such as an electrohydraulic unit or instrument air, or it can also be powered by an electronic motor which is usually accompanied by a small programmable logic controller (PLC) of some sort.
The third vital component to the sampling system is the container unit. In crude oil sampling, there are varying types of fluid that can range from high pressure, low gravity product to very low pressure, high gravity product. The designer of the system must be mindful of the type of fluid being sampled to ensure that the correct container type is selected for the application. For example, volatile liquids which contain entrained gases or a high Reid vapor pressure (RVP) would constitute using a container more sophisticated than your standard “milk jug” or atmospheric container. In most cases of sampling systems design we see two types of containers, which can be broken down into sub groups; the two main groups of containers would be atmospheric and constant pressure. Atmospheric containers can be used in a wide range of crude products. More specifically, the most accurate results when using these containers can be achieved when the product is of a stabilized state or does not seem to have any entrained gases that, if exposed to low or ambient pressures, may be released from the liquid and cause shrinkage. These types of containers are widely used in midstream applications for several reasons. One reason is, there are some types of containers that are transportable and small, hence they are easy to maneuver around the facility and two, generally speaking, the cost of these types of containers is less than the alternative in the same size. There are some very distinct advantages to the transportable variety of container. These containers are usually offered in smaller sizes specifically to accommodate easier transportation in and around the facility where they are to be used. Common sizes generally do not exceed five gallons in size and are accompanied by easy to handle designs to ensure the technicians are not strained when handling the containers. Some containers also come with certifications such as ASME certificates and DOT and Transport Canada approvals for safe and legal transportation on roadways. Although sizes may vary with transportable containers, there are some constants. All containers should have connection ports for a lab mixing system. These containers, while small in stature and good for transporting, still require a means of mixing the sample prior to use for analysis. This is usually done by attaching the container to what is commonly referred to as a lab mixer. The lab mixer is motor and pump driven and circulates the product through the containers until it is evenly mixed. Mixing is achieved inside the containers by another item which is a constant with these types of apparatuses known as the spray bar. The spray bar will have several perforations in it to help it act as a spraying nozzle which allow the fluid to disperse throughout the canister and therefore assist with mixing the product. After this is performed for a period of time, the product is ready for analysis. The other type of atmospheric container is similar to the transportable version but it is of a stationary variety. These containers are used when contracts call for a larger amount than a transportable container is feasible for, or when lab mixers are not readily accessible. In these type of scenarios, a mixing system is incorporated into the container and it is used as one unit. Once the entirety of the container is mixed, a “draw off” container, usually less than a gallon in size, is used to draw off of the large container and then transported to the lab. The second style of container is the constant pressure variety. These containers are also offered in different sizes usually ranging from 500 cc cylinder, up to 5 gallon cylinder as a common size, with varying standard sizes in between. There are distinct uses for constant pressure cylinder use in crude oil sampling applications; the most common of these is in high vapor pressure crude oils and condensate application. These types of crude oils, which are becoming more common with the advent of the shale play evolution, often require the use of these types of cylinder to ensure that the sample is maintained in the exact state it is received from the pipeline. Because of the bulky nature of constant pressure cylinders, it is common to incorporate a mixing apparatus into the cylinder with the larger styles having a power mixer that is used to agitate the main cylinder. A sample is then drawn off of this larger cylinder with a smaller similar type cylinder with these same features. This cylinder is then taken to the lab for analysis. This particular example is also commonly used in the sampling of light liquid hydrocarbons such as natural gas liquids (NGLs) and liquefied petroleum gas (LPG).
With basic components defined, the design of the sampling system can now be determined based on the application presented. In custody transfer, there are many considerations to be made for the design of a system. The customer most often will define in the contract or standard operating procedure (SOP) what type of system to use, or how they want the sampling system designed. Most often the considerations are of the following:
The first consideration of course is referring to the main line style sampler or the in-loop style sampler. The accuracy of either sampler has been a debate for years, and has been proven true for either style in one way or another. The theory behind the advantage to a main line system is that with proper main stream mixing the extractor can be inserted into the center half of the pipeline, less than five pipe diameters downstream of the mixing element and has the ability to capture the contents of the flowing stream without any impediment. This has proven to yield a highly accurate result in most cases and is a widely selected method in the United States for automatic sampling. The accuracy of this method has since increased by the addition of the automated purging system into the API 8.2 standard as a suggested method to increase accuracy in a sample system. While this method can increase accuracy throughout the sample system, it is particularly advantageous for the mainline system because of the long insertion required for most mainline samplers. The method by which this section of the sampler and sample line has been voided and prepared for the next batch is that once the previous batch is finished, the control unit would actuate the sampler a predetermined number of times until it can be determined that cross contamination does not exist. This portion of the sample is most often sent to the sump and not to the container in which the batch is supposed to go. This inherently minimizes the accuracy of the sample because the amount of sample still in the sample line, which may represent the last portion of the batch, would never make it to the container. Incorporating the purging method allows the user to displace the remaining portion of the sample into the container by sending a blast of concentrated inert gas or nitrogen through the sampler and the sample system, and thus into the container. This not only increases the accuracy of the sample, but also prepares the next batch for sampling and avoids any chance of cross contamination within the system.
There are also advantages that may be presented to using the in-loop style sampler. The one most commonly referred to advantage is that the section of the mainline pipeline that can be captured for sample is increased when using an in-loop system. Consider the fundamental difference between a main line or in-loop system. The mainline system has the extractor in the flowing stream and its point of capture or grabber is only a very small section of the main line pipe. The grab area of most samplers has an opening which is usually less than half of that of a fast loop take off probe. It is thought that a takeoff probe having a larger facing area to the flowing stream can pick up a larger profile of the stream therefore enhancing the accuracy of the downstream grab. While the surface area may increase, the in-loop system also incorporates additional variables which can also contribute to the decrease of its accuracy. The slip stream itself must maintain a flow rate close to equal that of the main line which means that in most scenarios, the incorporation of a pump into the system is necessary. Pumps have been shown to have the ability to “hide water” within the system itself if velocities are not adequate to continue the mixture through the bypass section of the system. If the pump is too slow, a bow wave will be created at the entrance of the probe, allowing the water to negotiate away from the entry of the probe, leading way to a lower water cut. Conversely, if the pump speed is too fast, it will create a suction of the probe entry, encouraging free water to enter the system at a higher rate, thus increasing the amount of water entering the system. This is a particularly grave issue when it comes to custody transfer and contract negotiations. Also, the addition of a pump invites yet another mechanical device and potential breakdown or maintenance item within the system. It is often recommended to oversize the pump for the application so that the pump does not operate at a high capacity, mitigating wear and maintenance stoppage. Secondly, even though the in-loop sampler may be designed so that the line pack area is minimized, it is still highly recommended to have a purge to void the small section of the sampler that still contains product. More so, it is still necessary to incorporate the automatic purge system to be able to void the sample system tubing of any remaining product from the batch.
The second consideration is referring to the setup of the containers and size of the system. Usually, the type of product will determine the type of container to be used, the most common being the atmospheric container or constant pressure container. Other considerations that may determine the type of container may be the size as well. Depending on the contract, or the frequency at which the sample is collected, the size of the container may change. Generally, no less than 500 cc would be taken in any container when being considered for custody transfer. As a standard, seven liters up to five gallons is the most common range for collection of a representative custody transfer crude oil sample. What also must be considered is the number of containers to be used in the system design. This number will be determined by the customer and is determined based upon the number of shippers or customers that are bringing in or receiving product from the custody transfer point. There should also be provision for additions to the system if necessary.
A few main factors in the system setup can spell the difference in performance and accuracy of the system. The idea of properly sized and downward sloping tubing should always be a factor in the container system design to ensure no water traps are created within the system. Proper switching mechanisms for container selection, and level indication devices are also important considerations for safety and overall functionality of the system. The most common of these devices are three or four-way automated solenoid selector valves, and high level switches or weigh scales respectively.
The final consideration when designing a sampling system is that no matter how well engineered the system, the matter of proving the system in the field is always a factor. The benefits of a well-established performance monitoring system can be useful in determining if an issue exists within the system prior to experiencing the issue during a custody transfer exchange. Situations such as sample controller system malfunction, seal failure within the extractor, sample pacing issues or inactivity during the batch flow are just a few examples of potential issues that should be tested out and remedied prior to engaging the system into a custody transfer scenario. Once the system is deemed acceptable for custody transfer, periodic testing should be conducted to reconfirm the system’s accuracy.
Ultimately, crude oil sampling is a mechanical function that relies on complex equipment to achieve highly accurate results. As discussed, any one of the components included in the design of a sampling system can be the determining factor in which high accuracy can be achieved, or diluted. For years these systems have been the cash register that allows shippers and customers to determine the value of the product in their pipelines and tanks across the planet, and with advances in sampling technology still taking place, higher accuracy is still achievable. It is absolutely imperative for the sampling system designer and user to stay up to date on all the latest techniques and methods of crude oil sampling. With millions of dollars changing hands daily in the oil and gas industry, sampling crude oil properly by automatic sampling remains an art and continues to be a widely accepted means of quality measurement determination.
Fish, David J. “Crude Oil Sampling.” Pipe Line Industry. April 1992. Print.
Jiskoot, Mark. The “Art” of Crude Oil Sampling. Cameron International Corp, 2011. Electronic.
American Petroleum Institute. Manual of Petroleum Measurement Standards Chapter 8.2, Standard Practice for Automatic Sampling of Petroleum and Petroleum Products. API, November 2016. Fourth Edition. Print.