API 18.2 recommendations improve level instrumentation

Updates to safety standards call for capable instrumentation to address safety and accuracy concerns.

By Lydia Miller September 30, 2019

Situations where money and merchandise change hands between two parties are often regulated by a third party to make sure the transaction is accurately made and recorded. When crude oil is purchased by a refiner or pipeline company, the custody transfer at the wellhead is regulated by the American Petroleum Institute’s API MPMS, Chapter 18.2 standard to ensure both parties understand and accept the transaction.

Prior to 2016, the methods for measuring and evaluating oil during custody transfer from a wellhead site were described in API MPMS, Chapter 18.1: Measurement Procedures for Crude Oil Gathered from Lease Tanks by Truck. This standard was published in 1990 and became widely recognized as the accepted method for this type of custody transfer. API MPMS, Chapter 18.1 was updated over the years before being supplanted in 2016 by API MPMS, Chapter 18.2: Custody Transfer of Crude Oil from Lease Tanks Using Alternative Measurement Methods, which is now the recognized standard and offers significant advances over its predecessor. At the time of its release, API issued this explanation:

“Industry standards developed by API support the industry goal of zero accidents,” said Lisa Salley, API’s vice president for global industry services. “This standard is a great example of what can be done when industry, regulators and all key stakeholders work together to achieve the common goal of improving safety for industry operations. This standard enables personnel to take measurements of crude oil from a lease tank without opening the hatch on the tank, thus protecting them from potentially hazardous vapors and gases.”

The new standard still allows for the old practices to be retained, but it offers mechanisms to use better methods able to make oil volume measurements more accurate, and the process of transferring oil much safer.

The old-fashioned method

Under API MPMS, Chapter 18.1, large production sites with frequent deliveries could be outfitted with a lease automatic custody transfer (LACT) skid. A LACT uses a sophisticated flowmeter to measure the volume of oil transferred along with other parameters, such as density, but it is not economical for smaller sites and volumes. The manual practices under API MPMS, Chapter 18.1, were designed to work anywhere, including wellhead sites that had effectively zero instrumentation. All the measurement methods could be carried out by a receiving truck driver using a kit of simple manual tools to take readings from the top of the tank. Here are the steps:

  1. The truck driver climbs to the top of the tank, opens a thief hatch and lowers a tape measure with a float to determine the oil level. This first measurement is called the opening gauge and establishes the baseline measurement.
  2. The next step is to lower a thermometer or temperature sensor to read the oil temperature which is used to support a density calculation. Depending on the oil depth, up to three readings should be taken at three depths due to potential temperature stratification. The density calculation uses an average of the three readings.
  3. A collector is dropped in to capture a sample. Specific gravity can be measured using a hydrometer corrected by the temperature. Some of the sample goes into a centrifuge for the “grind out” to determine how much water and solid material is mixed in. The measurements from a single sample are applied to the entire lot.
  4. If the driver is satisfied with the quality, the oil can be pumped into the truck. When the transfer is over, the driver takes a second level measurement (closing gauge) with the tape to calculate the volume transferred based on the tank’s dimensions.

All the necessary equipment was carried from site to site and nothing was automated. Everything had to be written down or manually entered into a computer. The ability to measure accurately was utterly dependent on each driver’s skill and finesse with the tape measure, which introduced significant variability. The U.S. Bureau of Land Management reported that typical manual tank gauging uncertainties range from 0.6% to 2.5%. Using a midpoint of 1.5% uncertainty and applying that to a well producing 600 barrels per day of oil at a sale price of $55 per barrel, this could result in an annual discrepancy of $180,000.

Another potential problem area is water accumulation in the storage tank. If the production separator is not working correctly or experiences an upset, water can be diverted to the oil tank or vice versa. If this problem is not detected, oil in the water tank will be sent out with the water hauler. If the oil-collecting driver doesn’t realize what has happened, water can be transferred along with oil. A manual level measurement from the tank roof using a tape cannot determine where an oil-water interface is in the tank.

The most dangerous aspect of the process is what can happen to the driver when opening the thief hatch. There is no specific number on how many drivers were greeted by an unwelcome blast of hydrocarbon fumes or toxic hydrogen sulfide expelled from the tank. The fortunate ones probably only suffered dizziness, fainting, headaches or nausea. Others weren’t so lucky. Between 2010 and 2014 the National Institute for Occupational Safety and Health identified nine fatalities of workers overcome during manual tank gauging and sampling operations. As mentioned earlier, this heart-breaking fact was one of the key drivers in the development of API MPMS, Chapter 18.2.

Using alternative methods

The new measurement methods in API MPMS, Chapter 18.2, include the ability to perform many of the relevant transfer measurements in either the trailer or truck zone or the transition zone rather than making them all in the tank zone. This allows the measurement taker to remain on the ground and away from many of the harmful vapors. The temperature can be monitored continuously during the transfer, and samples for “grinding” and density evaluation can be taken at any time from the pipe carrying the oil from tank to truck.

Volume transferred can be measured in one of two ways. First, a flowmeter can be used without the need for implementing a full LACT skid. API MPMS has chapters covering a variety of flowmeter technologies used for this purpose. Second, the opening and closing gauge measurements are still used, but with the measurement performed using a level instrument rather than manual measurement. Choosing the most suitable level instrument becomes a critical question. The application calls for several key capabilities:

  • A high-precision, fine-resolution continuous measurement from the top down
  • A simple mechanism requiring minimal maintenance
  • An ability to detect and locate an oil-water interface.

Meeting these three requirements with one technology narrows the field quickly: Radar instruments that mount from the top not only minimize the need for mechanical modifications to a tank, but they also have no moving parts. This approach is ideal since it provides the precision and resolution needed. API MPMS, Chapter 18.1, called for three consecutive manual readings to agree within 0.25 inches. The right radar level gauge can provide reliable readings with accuracy better than ±0.125 inches, meeting the first and second requirements.

The third requirement is the most difficult to meet, as few technologies other than guided-wave radar (GWR) (Figure 1) are capable of detecting and measuring the position of an oil-water interface. Magnetostrictive level instruments can be set up to capture an interface measurement, but effective operation depends on consistent densities of the liquids and free movement of the floats. If tar or other material from the oil accumulates on the rod it can interfere with movement, and the reliability of a reading will be lost. There is no way to tell this is happening from the reading data, short of the float freezing in one position.

A GWR probe can also accumulate buildup, but it takes a lot of material to interfere with the reading. Moreover, the nature of the echo curve can be monitored to indicate if material is building up, allowing appropriate maintenance action to be taken. Additionally, if there is an emulsion layer between the two liquid layers instead of a definite interface, the magnetostrictive device will always float at a point in the emulsion layer. If an operator relies on that measurement for the separation, some of the material identified as oil will be an emulsion of oil and water. Radar, on the other hand, will not measure well with an emulsion, indicating that separation is not complete.

Automated data collection

Absent a situation where a LACT skid is available, the manual actions of custody transfer under API MPMS, Chapter 18.1, did not lend themselves to automated data collection. Accurate record keeping depended on the fastidiousness of the driver either writing down the figures legibly or entering them without error into a laptop or tablet. API MPMS, Chapter 18.2, was developed specifically to allow for replacement of many of the manual measurement methods with the additional allowance for measurements to be made outside the more dangerous tank zone.

Using a GWR level gauge along with temperature transmitters and other electronic instrumentation provides the ability to tie readings directly to a data-gathering platform, greatly reducing the potential for errors. Additionally, the use of GWR for level measurement, instead of float-based technologies, can improve measurement reliability and help to monitor separation. The growing availability of WirelessHART instrumentation makes this easier to implement since these require no wired infrastructure for power or data transfer from the instruments. All the instruments necessary to perform a transfer operation under API MPMS, Chapter 18.2, are available as battery-operated wireless units, including GWR level gauges (Figure 2). Wired instruments are also available to perform all the necessary tasks.

These elements, working together, can produce a higher level of accuracy and reliability while avoiding potentially lethal safety concerns. API MPMS, Chapter 18.2, shows how users can put these techniques to work to produce the desired results in a safe and repeatable manner with reduced requirements for onsite labor.


Author Bio: Lydia Miller is a product manager with Emerson Automation Solutions, working with Rosemount level products, with a focus on radar and ultrasonic instruments and level switches. She joined the company in 2011 and has additional work experience with air-to-air energy recovery for process industries and HVAC applications. Lydia has a bachelor’s degree in mechanical engineering and English from the University of Minnesota.