September 2019

Process Control and Instrumentation

Automatic gauging addresses tank farm challenges

Tank farms are often found on the premises of large refineries and petrochemical plants, or nearby.

Mallon, D., Endress + Hauser GmbH + Co.

Tank farms are often found on the premises of large refineries and petrochemical plants, or nearby. Operators of tank farms face several challenges, not the least of which are outmoded methods of measuring tank volume (FIG. 1). Older tank farms often work with outdated technologies, which can put their employees at risk and cause spills, leading to fines. Even newer tank farms can face the same problems if they do not have the right tank volume and level measuring systems in place. This article will examine these challenges, and then describe how automatic tank gauging addresses these issues.

Tank farm challenges

Safety concerns are priorities for tank farms. Over the last several years, more companies are mandating that their workers stay off tank roofs and not open tanks, to eliminate safety risks due to unsafe working conditions and/or human error.

FIG. 1. Oil tank farms face challenges that automatic tank gauging can help solve.
FIG. 1. Oil tank farms face challenges that automatic tank gauging can help solve.

Tank gauging is usually done by a person working alone, who lifts open the hatch of a storage tank to manually measure the liquid level inside the tank. Gases escaping from the open hatch can overcome personnel and can be fatal. These incidents have led to lawsuits against tank farm companies.

A major portion of the problem involves vapor recovery units that do not allow gases to vent from tanks into the atmosphere. In 2012, the US Environmental Protection Agency (EPA) mandated that all new tanks be outfitted with closed-loop systems. The rule inadvertently increased the danger to workers who manually gauge these tanks (FIG. 2), as closed-loop systems expose these workers to vapors that exist under greater pressure and gush out on release. Measuring tank levels with a stick can result in inaccurate readings, as well. This method will be covered in detail later in this article.

FIG. 2. Tank roof work poses a height danger for workers. Also, manually measuring tank levels can result in inaccurate readings.
FIG. 2. Tank roof work poses a height danger for workers. Also, manually measuring tank levels can result in inaccurate readings.

Complying with new standards from the US Occupational Safety and Health Administration (OSHA), the American National Standards Institute (ANSI), the American Petroleum Institute (API), the National Fire Protection Association (NFPA), and other US state and federal regulations can be challenging. Many plants and tank farms were built prior to the implementation of present safety standards. The legacy control systems and instrumentation, primarily the tank level instrumentation, may be obsolete or require updating.

In most legacy systems, the level instruments were selected long ago to provide measurement data to the automation system. These instruments were usually not chosen based on the financial impact the measurements could have on profits. Modern level instruments fulfill both roles, providing both precise control of operations and accurate financial accounting.

The API 2350 standard provides tank operators with best practices for preventing overfill events in petroleum storage tanks. The latest version of API 2350 is scheduled for publication at the end of the year. This standard is covered in more detail later in the article. Although API 2350 is not a regulation, tank farms are under pressure to conform to the standard.

Economics of inaccuracy

Measuring tank levels with a stick is not only dangerous, but it is also inaccurate. People can read or interpret the measurement differently. The same person does not always read the same tank, and the level indication is accurate only to the nearest inch or so, at best.

After a reading is made, it must be written down and taken back to the control room to be entered into a tank inventory database. Since the person probably has a dozen or more tanks to measure, the reading may not be entered for hours or even days, so it may not be up to date when it is finally entered. At best, manual readings provide only a rough estimate of tank levels.

Legacy level instrumentation can pose challenges, as well. For example, consider a tank being filled or loaded with oil. It has a capacity of 30,000 gal, and the system will stop filling when the vessel gets to 80%–85% of capacity. The process engineer decided that an accuracy of 2% was acceptable for control purposes because, if the tank reached 82% or 87% of capacity, respectively, it still would not pose an overspill risk. With a 2% accurate instrument and the control mechanism programmed to stop at 85%, an error of up to 600 gal could occur (2% × 30,000 gal = 600 gal).

If a customer wants to load out 5,000 gal of oil, how accurate should a measurement instrument be to ensure that the customer is getting the entire 5,000 gal? With instrumentation that is only 2% accurate, the error would be 100 gal.

If the transaction will happen only twice a week, the unaccounted loss could be as high as 200 gal. If this transaction occurs 100 times a week (a more typical frequency), then there could be an unaccounted loss of up to 10,000 gal. If the product sells for $2.50/gal, then an unaccounted financial loss of up to $25,000/wk is possible. Conversely, if a 0.25% accurate instrument is chosen, then the unaccounted loss can be reduced to 750 gal, or $1,875. Accurate instrumentation can save up to $22,125/wk.

Note that the unaccounted loss can be in favor of the tank farm (i.e., the tank farm is loading out 100 gal less per load). Customers sometimes check to ensure they are getting the proper amount and will raise concerns if they do not get the exact 5,000 gal. Conversely, if the tank farm is loading out more than 5,000 gal, it probably will not hear anything from the customer, but it will lose potential profits.

Measuring level does not necessarily provide volume. In perfectly round tanks, the volumetric calculation is simple, but many tanks are not perfectly cylindrical. Tanks can flex when being filled, expand with higher temperatures or have internal obstacles. When filling, oil foaming and mechanical waves can occur. To accurately determine volume, especially for tank tables that show volume at various levels, it is important to allow settling times for waves and to consider temperature compensation and other calculations.

API 2350 requires accurate level measurements

Reducing the risk of an overfill is the main objective of API 2350. Different levels of risk are associated with each tank because the filling rates are different, and sensors have varying reliabilities, among other factors. The likelihood of an overfill event requires tank-specific information such as filling rate, installed instrumentation and other data. If a tank is at a high risk of overfill, it will need additional level instrumentation (FIG. 3).

FIG. 3. API 2350 defines a properly instrumented and engineered tank. For example, this tank has an independent overfill protection system, a high-high point level switch and an emergency shutdown valve.
FIG. 3. API 2350 defines a properly instrumented and engineered tank. For example, this tank has an independent overfill protection system, a high-high point level switch and an emergency shutdown valve.

Risk assessment determines the level of risk, and properly addressing those risks makes up a risk management process. Risk assessment determines the likelihood and consequence of harmful events, but just knowing the risk level does not help unless those risks can be reduced to an acceptable level, which requires risk management. Proper application of this methodology reduces risk by implementing corrective actions.

API 2350 specifically states that a risk assessment must be implemented and maintained, but it does not specify how to do this because risk assessment methods and techniques often vary from one plant or facility to the next.

A risk analysis process incorporates the results of data collection, organization, analysis and assessment. For example, specific tank information (such as filling rate, previous spills or near misses, operator reliability, implementation of an automatic overfill prevention system and gauging system reliability) can be compiled and included in a database. In addition to data on size and frequency of events, the severity of various failure modes is also included. For example, the severity of an overfill or spill event may be influenced by the filling rate, the hazard associated with the liquid, the proximity to populated areas, containment capabilities and the likelihood of early awareness.

FIG. 4. This proprietary dual-flange level switch<sup>b</sup> provides both high and high-high level indications.
FIG. 4. This proprietary dual-flange level switchb provides both high and high-high level indications.

A complete tank instrumentation system is shown in FIG. 3, but, at the very least, a tank should have both a high level and a high-high level switch (FIG. 4) installed to prevent spills. Complying with API 2350 often requires new instrumentation—from just high and high-high level switches on tanks with low overfill risk—to complete overfill protection systems for tanks at high risk.

Integrating the instrumentation

In a tank farm with 100 or more storage and transfer tanks, the number of new level transmitters, flowmeters, level switches, and pressure and temperature sensors needed to instrument the tanks can be considerable. Since all of this instrumentation must be connected to the tank farm’s control and monitoring systems, a systematic approach addressing the most critical risks first is paramount.

In addition, the system must perform a considerable number of calculations to convert level signals to volume. This measurement means accommodating odd-shaped tanks, performing temperature compensation, using tank tables and taking other factors into account.

If a new control system is being installed simultaneously as the instrumentation, the problem could be simpler because the new instrumentation and control systems can be designed to work together. If the tank farm is trying to keep its legacy control system, challenges can arise that would need to be addressed.

FIG. 5. A tank scanner, such as the proprietary onec shown, provides tank correction tables for innage and ullage, temperature and volume corrections, density tables and mass/weight calculations.
FIG. 5. A tank scanner, such as the proprietary onec shown, provides tank correction tables for innage and ullage, temperature and volume corrections, density tables and mass/weight calculations.

First, all the new instruments must be wired to the control system, which probably will require new or expanded input/output (I/O) capabilities. Next, the system must be programmed to perform the level-to-volume calculations, which can require a substantial amount of custom coding.

One way to avoid this problem is to put the necessary calculations “at the edge.” This term means at the tank itself with automatic tank gauging (ATG) hardware and software. The initial ATG device is a tank scanner. Several instrument vendors offer these types of devices, which typically contain tank correction tables to perform the necessary calculations for providing the control system with the correct volume (FIG. 5).

Since the tank scanner inputs all the necessary information from flow, level, temperature and other instruments, it can also provide this information to the control system via a single digital data link. This eliminates the need to wire every instrument back to the control system individually, with a corresponding discrete or analog input, or both, for each.

Dealing with data

With 100 or more tanks, the system must keep track of a large amount of data. This includes which tanks contain what products, how much is in each tank, how long it has been there, and how frequently the vessel is used.

It is possible to enter all this data into a process historian, spreadsheet or database, and then write algorithms so that operations can determine the following:

  • How much of product X do we have?
  • Where is it?
  • Do we have room to store 500,000 gal of product Y?
  • Where can we put it?
  • Flowrate with line size?
FIG. 6. Inventory control software can take data from scanners and store it for easy access by tank farm operators. This softwarec can store data on 400 storage tanks.
FIG. 6. Inventory control software can take data from scanners and store it for easy access by tank farm operators. This softwarec can store data on 400 storage tanks.

Instrument manufacturers have come to the rescue with ATG inventory control software (FIG. 6) that operates on a separate PC or in the cloud. For tank farms with legacy control systems, this is an added benefit because it offloads the entire inventory control problem to an external system, which can communicate with the legacy system via a digital data link.

In addition to the basic ATG hardware and inventory control software, instrument vendors typically offer add-on packages that use the database in the cloud to perform various functions, including:

  • Ship unloading—This process allows for the comprehensive planning and safe operation of liquid loading and offloading to and from tanks.
  • Custody transfer—An accurate measuring system, which complies with local custody transfer law, is required to monitor custody transfers. With an automated and certified tank gauging system, operators can create the legally required documentation.
  • Truck-loading operations—This system controls truck-loading operations, including truck and personnel validations, permissive handling, precise reading of loading quantities and successful storage of completed transactions in the system.
  • Enterprise resource planning (ERP) integration—This process communicates relevant data such as orders and transactions, as well as truck and driver details from or to a central software-based ERP system.
  • Queuing optimization—An important aspect of running a tank farm is the transfer and offloading time. All loading deliveries must be executed as quickly as possible, and a queuing system is an effective tool to do so. For example, a truck queuing system allows the user to plan truck-loading operations to optimize the throughput of trucks at the tank farm.

Takeaway

With ATG tank scanners and inventory control software, the entire process of updating a tank farm can be simplified, and the resulting system can provide excellent inventory control and improve operations—thus decreasing the amount of pressure that tank farm operators face. HP

NOTES

a Endress+Hauser’s Micropilot NMR84 radar measurement device

b Endress+Hauser’s FTL51 dual-flange level switch

c Endress+Hauser’s TankVision tank scanner

The Author

Related Articles

From the Archive

Comments

Comments

{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}