The ownership of crude oil changes frequently during its transport from the reservoir to the final market destination. With each change in ownership, buyers and sellers engage in a transfer of custody and must rely on the measurement of the asset. The accuracy of the metering measurements is critical.
During its spring lecture series, the SPE Gulf Coast Section’s Projects, Facilities, and Construction study group presented sessions on experts’ discussions of the processes involved in ensuring accuracy in measurement, meter selection and proving, multiphase flow measurement, and comparisons of gas measurement technologies.
Del Major, a liquid hydrocarbon measurement adviser at Shell, analyzed the best practices in selecting and maintaining fit-for-purpose meters and for mitigating bias in the measurement of hydrocarbons in a reservoir.
Bias is an inaccuracy in measurement. It is not the same as uncertainty in measurement, which accounts for a range of potential values close to the true value. Bias describes a consistent deviation. Major said the key assumption in measuring is that the risk of inaccuracy is random and equally distributed around the true value, but a consistent deviation represents an error in the system.
Major compared a meter to a thermostat. If a thermostat shows a temperature of 80°F with a margin of error of 3°F above or below the reading, the margin would represent uncertainty in the listed temperature. However, if it is known that the thermostat reading is always 3°F below the true temperature, it would indicate bias.
Shell classifies the meters into four categories as follows:
Each category has a given level of uncertainty ranging from 0.25% for fiscal sales meters to 20% for environmental and emissions trading meters. Major said all single-phase metering applications have levels of uncertainty because of their exposure to processing. Some meters are more accurate than others because of prevailing flow/fluid properties and conditions. Still, companies should aim to reduce uncertainty within metering systems.
“Uncertainty is a function of the quality of the equipment and how it is used, so when we talk about uncertainty, it’s really the inverse of accuracy. The lowest uncertainty provides the highest confidence that our measurements surround the true value. If we see there is a shift and it is biased, then that is mismeasurement and the one thing that we’ve got to do is eliminate any bias,” Major said.
To add value to the life cycle of a reservoir, Major said companies should consider single-phase devices that are insensitive to changing fluid properties such as density, viscosity, pressure, and temperature. A good device should also require little or no maintenance or recalibration, and should have a high turndown ratio.
The turndown ratio is a measure of the usable range of an instrument, or the high end of a measurement range compared with the low end: Turndown ratio=maximum flow measurement/minimum flow measurement.
Tony Petitto, a technical sales manager at FMC Technologies, discussed the basic operating principles and application ranges of metering technologies. He said that reducing measurement uncertainty and eliminating bias is the main goal of accurate flow measurement, because the smallest error may lead to significant financial losses for a company. The key factor to obtaining accurate, consistent measurements with low uncertainty and no bias is the selection of the right meter technology for a given application. Companies can decrease long-term earnings by not investing in the correct systems for their needs.
“If you try to save some money by not putting in the right [meter], or you don’t put in all the equipment that is really needed, the costs of a bad measurement can be a lot more in the long run than whatever you spend in buying and maintaining the equipment,” Petitto said.
In-situ proving of a meter at given temperatures and pressures that replicate actual operating conditions is the preferred way to verify accuracy and reduce uncertainty in the system, Petitto said. The method compensates for variations in meter performance caused by installation effects and changes in operating conditions.
A well-designed custody transfer system should have the right meter for the application, and the decision should be based on product characteristics and process conditions, Petitto said.
The system should also have the right number and size of meter runs, flow control and balancing valves, double block and bleed valves, and a prover for the application. A prover helps to establish reference points under operating conditions, verify repeatability/linearity of a meter, account for changes in accuracy, and satisfy the contractual requirements of a custody transfer.
The measurement of oil, gas, and water coming out of a reservoir is a difficult task for operators. When a test separator is too costly, big, or complex to be installed on site, the task becomes more difficult to perform. When an application does not allow for the installation of a test separator, fluid measurements must be determined with an inline multiphase flowmeter.
Lars Farestvedt discussed the current state of multiphase meters and wet gas flowmeters in the oil and gas industry, including the ways in which they can help solve difficult problems often faced in measuring fluids. He is the general manager of Multi Phase Meters at FMC Technologies.
Over the past 2 decades, multiphase and wet gas flowmeters have gone from a prototype to a mature technology widely used around the world. Although “several thousands” of these meters are estimated to be in operation today, Farestvedt said they have not reached a stage where they could be considered as “plug and play” technology.
“When you put [multiphase meters] in the field, they need to be set up correctly,” he said. “They need to be handled correctly for them to work. And that’s true really for any type of technology, even for single-phase meters. That’s something to be considered.”
Farestvedt raised several points with multiphase and wet gas flowmeters that require further attention from operators and vendors when choosing meters. Operators must look at the price of various meters, which can vary significantly because of geographic and weather-related conditions surrounding a reservoir. A test separator in west Texas, for example, may cost much less than a separator in the Alaska North Slope.
Additionally, Farestvedt said it is important for operators to understand the meter applications available to them. “There is a wide range of applications for high gas void fraction, low gas void fraction, low and high water/liquid ratio, switching areas between multiphase and wet gas. There are a lot of challenges multiphase meters need to handle, and you need to consider what the application is and whether the meter that you’re looking at can cover the range that you need,” he said.
There is a limited number of manufacturers of multiphase and wet gas flowmeters, and the technology is diverse. Farestvedt said that specifications for uncertainty and performance, and the influences of and sensitivities to external parameters, such as fluid properties, may differ among systems. Operating envelopes and the presentation of test results are also varied.
Farestvedt said some guidance is required to guarantee a uniform approach in testing and specifications. “There are many things that differentiate the technologies,” he said. “Making standards to cover multiphase meters is not very easy. For an orifice fitting, that’s a lot easier because it has to have certain dimensions. For turbine meters, they’re fairly much the same way. Those types of technologies are standard, but here [in multiphase meters] you have a whole bunch of different technologies. It is not too early to make guidelines and standards on how to specify, test, implement, and use this technology.”
Dan Hackett discussed four types of metering systems: orifice, turbine, ultrasonic, and the Coriolis flowmeter used in the custody transfers of oil and gas. He is the business development director of ultrasonics at the Daniel Measurement and Control division of Emerson Process Management.
Orifice meters are the primary meters used in custody transfers of liquid and gas. They are well documented in standards, widely accepted, and the requirements for their use and maintenance are known in the industry. Orifice meters are relatively cheap to purchase and install, and have no moving parts in the flow stream. When built to standard requirements, they do not require calibration beyond the confirmation of mechanical tolerances or periodic calibrations.
However, orifice meters have a low rangeability—the maximum controllable flow/minimum controllable flow ratio—with a single readout and a high pressure loss at a given flow rate, particularly at lower beta ratios. The beta ratio is expressed as orifice plate bore/inside pipe diameter. The instruments are more sensitive to flow disturbances at higher beta ratios than other meters, and the flow pattern in the meter prevents its self-cleaning.
Despite these disadvantages, Hackett said that orifice meters are still reliable for use in liquid and gas custody transfers. “The old saying is ‘nobody ever got fired for putting an orifice meter in.’ That’s still true. It’s a good device and it has its place,” he said.
Similar to orifice meters, turbine meters are widely used in both liquid and gas measurements. Hackett said they provide good accuracy over the full linear range of the meter. The electronic output is available directly at a high resolution rate, which makes proving of the meter possible in a short period of time. Turbine meters cost slightly more than orifice meters, but the cost to install the total turbine meter station is still fairly low because of the high flow rate for its given line size.
Also, the turbine meters handle normal flow conditions well despite their pressure and temperature limits, and they have an excellent rangeability on gas meters at high pressures, Hackett said.
Among their disadvantages, turbine meters require proving with throughput to establish their most accurate use, and while the rangeability at high pressures is good, the rangeability at low pressures is about the same as with other gas meters. They require an upstream flow pattern that is nonswirling, thus necessitating the use of a long inlet pipe or straightening vanes.
Being the most fragile of meters, turbine meters are more acceptable for liquid measurement than for gas, Hackett said.
“Even though [a turbine meter has] good accuracy and great linearity, and high-resolution pulse output, you’ve still got to put some kind of strainer or flow conditioner in there. You’ve basically got to keep the bearings and the blades from getting torn up by stuff that’s running down there. These other technologies may require a flow conditioner, but not a strainer,” he said.
Ultrasonic meters do not experience any pressure drop, since they have the same diameter as the adjacent piping, and the high-frequency pulse rate of output minimizes errors from the effects of pulsation and fluctuating flow. Hackett said their installation is simple and cheap despite a high initial capital cost. They have a high rangeability and no moving parts that make contact with the flowing fluid.
On the downside, unlike an orifice or turbine meter, the flow profile of an ultrasonic meter must be fully developed to determine an average viscosity, either from a single path or from a reflection unit.
“Any ultrasonic meter is going to be some kind of arrangement of paths, whether they’re crossing or a mix-or-match. They’re only taking a certain number of samples. We have to take three points, or four points, or five points and integrate those velocity measurements into a volume number. If I don’t have a repeatable profile, then the weighting factors or the integration algorithm to take velocity to volume may not stand up,” Hackett said.
Coriolis meters have narrow tubes, which may require the use of a coarse filter to prevent debris from clogging them. Hackett said they have no problems with moving parts except for a vibrating element that may be troublesome, depending on the stresses and temperature gradients involved in the measurement.