LIQUID MEASUREMENT STATION DESIGN Class # 2230.

1 Bruce Byers Systems Sales Director North America FMC Technologies, Inc. 1803 Gears Road Houston, Texas USA Introduction Liquid Measurement Stations are necessitated by agreements between petroleum buyers, sellers and transporters along with appropriate customs and or governmental authorities. These agreements outline how the fluid is to be measured and how the results will be traceable to recognized standards. In the case of common carrier pipelines, the pipeline is entrusted with the transport of their customers' fluid, thus loss control by use of accurate liquid measurement stations is essential. It is important to note that everyone involved has an interest in the true net volume or associated mass. In addition to meeting the requirements for measurement, stations must meet numerous safety and construction codes and standards, as the fluids are normally hazardous. Operation of the measurement station must be relatively simple including a user-friendly, man-machine interface that effectively manages the flow measurement and control requirements. The task of the station or system designer is to transform these requirements into engineering specifications, drawings, and bills of material for procurement, manufacture, test certification and delivery to the end user of a multitude of components specifically selected and integrated to work together to meet the requirements of the measurement agreement and applicable codes. This paper will discuss the various topics the designer must address and the methodology he must use to produce a satisfactory system. System Considerations High performance liquid meter stations offer state of the art solutions to high accuracy liquid measurement requirements; however, for achievement of ultimate performance it is important that the basic measurement elements be surrounded by carefully selected ancillary components and all items be integrated into an accurate well designed, easy to operate system. This necessitates a systems approach to the procurement of meter stations usually with a turnkey contract to a qualifed, experienced meter and system supplier. Overall uncertainty of the order of 0.25% or better is achievable with properly designed systems and under some circumstances the accuracy may approach 0.1%. Fluid Properties In order to optimize the system cost and performance and to avoid changes as the system is designed, it is desirable to have complete fluid and system environmental information (reference Table 1 - below) at the outset. At times, it is impossible for the user to define all of the parameters. If reasons why a particular parameter are not available or clear to the designer, he can allow flexibility in his component choices. An example: if viscosity is unknown, a meter with a wide viscosity range can be selected. It is also important that the measurement station designer be fully informed of the system configuration upstream and downstream of the station. Pressure drop and flow control requirements must be fully understood for properly designing a system. Environmental Considerations Many measurement stations are subjected to marine environments, and equipment must be selected with this in mind as well as any extremes of ambient temperature that will be encountered. Space available for the system and shipping size and weight limitations must also be incorporated into the system design. Protection of the surrounding environment from toxic spills and vapors is also a major design consideration. Drains, vents and drip pans must be specified when required.

299

DESIGN CODE SPECIFICATIONS

I. General design standards-measurement accuracy requirements API-Manual of petroleum measurement standards Chapter 4 - ProvingSystems Chapter 5 - Section 2 - Displacement Meters Chapter 5 - Section 3 -Turbine Meters Chapter 5 - Section 4 - Instrumentation or Accessory Equipment Chapter 5 Section 8 Ultrasonic Flow Meters Chapter 6 Section 5 - Metering Systems for Loading and UnloadingMarineBulkCarriers Chapter 12 - Calculation of Petroleum Quantities Chapter 13 - Statistical Aspects of Measuring and Sampling Requirements of national and local customs, weights and measures, and government authorities. Chapter 17 - Marine Measurement Chapter 21.2 Flow Measurement Electronic Liquid Measurement II. Pressure retention safety and construction codes ANSI/ASME B31.3 Chemical Plant and Petroleum Refinery Piping ANSI/ASME B31.4 Liquid Petroleum Transportation Piping Systems ASTM material specifications ASME boiler and pressure vessel codes NACE-National Association of Corrosion Engineers specifications III. Electrical safety codes National Electric Code-NFPA-70 Electrical codes of ship certification agents (if shipboard equipment) Local or regional electrical codes

Table 1

IV.

Structural codes AWS D1.1 - Structural Welding Code AISC-Steel Construction Code Structural codes of ship certification agents (if shipboard equipment)

Table 2 Codes and Standards Selection of appropriate and applicable standards and codes is essential in order that the system be accurate, safe, convenient to operate, and reliable. The standards most often used are listed by category in Table 2. Meter Selection Selection of meters is primarily influenced by flow rate, rangeability (turndown), fluid viscosity, pressure drop, wax content, and corrosive properties of the fluid stream. For multi-fluid streams, the worst-case fluid must be considered. Normally the selection is narrowed down to a positive displacement device vs. turbine meter vs. a mass meter vs. an ultrasonic meter. There are two types of turbine meters now available, the conventional rotor meter, and the helical rotor or multi-viscosity turbine meter. Each has its advantages and disadvantages for an application and they must be carefully considered. Please refer to Figure 1 and also References 1 and 2.

300

Figure 1 PD and Turbine Meter Selection Guide One important item to note is that most meter ratings vary with density and viscosity of the fluids to be measured. Mass meters are effective for measurement of highly corrosive and rapidly changing fluids. Ultrasonic meters may be favored for high flow rates. Mass and ultrasonic meters must be equipped with provisions for proving like any other custody transfer meter. Practical understanding of and provision for zeroing these meters is important. In the case of tanker loading and/or unloading facility design and FSO or FPSO applications, the station flow rate is determined by tanker capacity divided by desired maximum loading time. For very large tankers load times of 20 hours maximum are often used. This would result in a 50,000 BBL/HR system flow rate for loading a 1,000,000 BBL tanker in 20 hours. These are high intermittent flow applications. Obviously, continuously flowing lines more typical of pipelines or refinery output would have much lower hourly rates but comparable total volume throughput over the long term. A balance between rangeability, prover economics, maintainability, and overall system economics determines the size and quantity of meters for a particular station. For additional details see Reference 3. A single large meter requires a large prover; which results in limited rangeability, and would require complete shutdown of the station if any critical component in the meter run malfunctions. A large number of smaller meters would result in a much smaller prover, wide rangeability, and standby capacity in the event of a malfunctioning component. However a station with too many small meters has a large number of valves, and therefore the cost of pressure and temperature instrumentation, which is fixed per meter run, becomes high.

Typical Turbine M eters FlowRates (Based on liquid of 0.9 sp.gr., and 33 SSU(2CTKS) viscosity) US-GPM "K" Factor Imperial GPM Nominal Extended (pulses/ Nominal Extended M eter M inimum M ax. Rated M axim um volume) M inimum M ax. Rated M axim um Size Linear @6 PSID @8 PSIG 5% Linear @6 PSIG @8 PSIG 4" 140 1,000 1,280 25.0 120 900 1,535 LF6" 250 1,800 2,300 25.0 200 1,500 1,915 6" 360 2,800 3,500 25.0 300 2,500 2,915 8" 700 5,300 6,500 12.5 550 4,500 5,410 10" 1,000 8.500 11,000 12.5 950 7,000 9.160 12" 1,700 13,000 15,000 8.3 1,200 10,000 12,490 16" 2,500 19,000 24,000 2.5 2,000 16,000 19,965 18" 3,100 25,000 29,500 2.5 2,800 21,000 24,565 20" 4,000 30,000 35,000 2.5 3,500 25,000 29,140 24" 5,500 42,000 50,400 1.8 4,500 34,000 41,965 Barrels Per Hour Cubic M eters Per Hour 4" 200 1,500 1,780 1,060 30 250 290 LF6" 350 2,575 3,200 1,050 55 400 520 6" 500 4,000 5,000 1,050 80 640 780 8" 1,000 7,500 9,200 525 160 1,200 1,460 10" 1,500 12,000 16,000 525 250 1,900 2,450 12" 2,500 18,000 21,500 265 400 2,900 3,400 16" 3,500 27,000 33,500 105 560 4,300 5,400 18" 4,500 35,000 42,000 106 715 5,800 6,600 20" 5,700 42,000 50,000 105 900 6,600 7,800 24" 8,000 60,000 72,000 75 1,250 9,500 11,300

Typical P.D. Meters Size & CONTINUOUS Model USGPM BPH Carbon Steel C2 125 E3 420 E4 420 F4 600 G6 1,000 H8 2,285 JA10 3,500 JB10 4,700 K12 7,200 M16 12,500 L/min 475 1,600 1,600 2,250 3,750 365 550 740 1,140 2,000 Maximum Flow Rates INTERMITTENT 3 MH USGPM L/min 150 500 600 720 1,200 -----570 1,900 2,250 2,750 4,600 -----ALL IRON or LPG USGPM BPH L/min 100 315 315 450 750 1,200 375 1,200 1,200 1,700 2,800 1,720 2,600 3,525 5,400 9,375

Table 4

Typical Helical Turbine Meters Meter Min Flow 3 Size BPH M /HR 3 in 90 14 4 in 190 30 6 in 400 64 8 in 750 119 10 in 1250 199 12 in 1900 302 16 in 2700 429

BPH 900 1900 4000 7500 12500 19000 27000

Max Flow 3 M /HR 140 300 640 1190 1990 3020 4290

Table 3

301

Table 5

Most system designers recommend a design with a complete spare meter run. This allows a great deal of flexibility in that full system capacity is not compromised should one meter or critical component in a meter run become inoperable. It also allows a method of increasing total station flow rate either temporarily or permanently and allows use of all runs at a rate below their maximum, which results in lower pressure drops, and longer meter life. As a result of the above considerations, the normal multi-meter run station usually has a minimum of two meter runs and a maximum of 5 or 6 meter runs. The exact number is usually determined based on dividing the total station flow rate desired by a number from 2 to 6 until the rate per meter run is close to the maximum rate of a standard size meter as obtained from a typical manufacturer's meter size vs. capacity chart, reference Tables 3, 4, 5, 6 and 7. Note that typically these charts cover a certain density and viscosity range and they may need to be corrected before the meter size is selected. Selection of meter type, size, and quantity is usually the first and most important step of station design. Table 6
Typical Mass Meters Flow Rates 2 Pounds/Minute(kg/min) Gallons/Minute (Liters/Min) Meter Model Range 1 Range 2 Range 3 Range 1 Range 2 S25 0.8-8.0 1.6-16 24-24 0.09-0.9 0.19-1.9 (0.35-3.5) (0.7-7.0) (1.05-10.5) (0.35-3.5) (0.7-7.0) S50 5.0-50 10.0-100 15.0-150 0.6-6.0 12-12 (22-22) (4.4-44) (6.6-66) (2.2-22) (4.4-44) S100 20.0-200 40.0-400 60.0-600 2.4-24 4.8-48 (9.1-91) (18-180) (27-270) (9.1-91) (18-180) S200 75.0-750 150-1,500 225-2,250 9.0-90 18.0-180 Meters supplied from factory set at Range 2. Other ranges are field programmable. 2 Based on a liquid of 8.3 lb/gal. 1 kg/HR, e.g. water at standard conditions.

Table 7
Ultra Sonic Flow Meters
Range 3 0.29-2.9 (1.05-10.5) 1.8-18 (6.6-66) 7.2-72 (27-270)

Size Maximum Inches (mm) BPH

6 (150) 8 (200) 10 (250) 12 (300) 16 (400) 20 (500) 4,500 8,000 12,500 19,000 28,000 43,000

Flow3Rate m /h
720 1,270 1,990 3,020 4,450 6,835

Flow Conditioners When turbine or other inferential meters are used, special upstream and downstream flow conditioners are required to prevent fluid swirl and non-uniform velocity profiles. Also bores of upstream and downstream sections must be properly aligned and concentrically equal within the specified tolerances. Prover Sizing and Selection of Prover Type The next most important step is the selection of the prover or proving philosophy. The prover enables the meter to produce results close to the truth, i.e. an internationally recognized standard. The prover is a device that enables on line absolute volumetric calibration of the meters with traceability to The National Institute of Standards and Technology. Provers are calibrated by a "Water Draw" procedure to establish base volume at a reference temperature and pressure, usually 60F and 0 PSIG, using NIST calibrated measures. They are then used, "in situ on the fly" to calibrate the meters using the exact fluid being measured. For large high capacity systems each meter is proved at each flow rate at which it is used for each tender, batch or tanker load. Pipeline meters are usually proved periodically (perhaps daily or weekly) depending on flow rate changes, temperature, and pressure changes or fluid changes. It is recommended that inferential meters, i.e. Turbine meters, be proved more frequently than direct measuring meters such as PD meters. This is because it is possible for inferential meters to become grossly inaccurate due to various outside influences. Todays advanced control systems can be programmed to initiate proving based on changes of multiple parameters, such as flow rate, viscosity, density, time or throughput (Volume or Mass) since last proving. Provers may be bi-directional, uni-directional, piston type, or small volume. Bi-directional provers are the most widely used and tend to be the best choice for most applications. Spheres are available in a variety of elastomers. For use at fluid temperatures below 20F or above 200F a piston type prover is usually favored. Where space is

302

very limited and time or fluid volume to accomplish a series of proves is to be kept to a minimum, small volume provers are ideal. Small volume provers also are favored for chemicals, which may attack the available elastomers for spheres. When a dedicated prover is not justified for a meter station a trailer mounted prover is normally recommended. Figure 2 shows a typical bi-directional prover.

Figure 2 - Typical Bidirectional Folded Loop Meter Prover Prover design, sizing and calibration are standardized by API. A prover is sized at either 0.5% of maximum hourly flow rate (i.e. a 20,000 BPH bi-directional prover would be 100 barrels between switches) or for a minimum number of pulses from the meter, normally 10,000 pulses. Some applications require that both criteria be met. Pulse interpolation prover counters can be used to increase resolution of small provers or for meters with low pulse resolution. Minimum distance between switches is then controlled by the repeatability of the displacer detector switches. For more information on detailed prover design see Reference 4. Normally the prover for a measurement station is designed for the maximum flow rate per meter run, thus a 50,000 BPH station with four meter runs would have a 12,500 BPH prover. The maximum prover flow rate may or may not be the maximum capacity of the meter itself. For maximum system expansion capability it is usually recommended that the prover have the capacity to prove at the maximum meter rate. Block Valves Of utmost importance for accurate proving are properly selected valves. Reference Typical System P&ID Figure 3. When the meter run to be proved is aligned with the prover, no fluid must be added or allowed to escape from between the meter and the prover. This can only be assured by using double block and bleed high integrity valves in this portion of the system. In addition, since proving is done rather frequently these valves must endure many open and close cycles with positive sealing after each closure. Some systems require devices to monitor the valve integrity automatically. Any drain or vent valves in this portion of the system must also have means of monitoring that they are completely sealed. Provers may be arranged upstream or downstream of the meters. When the prover is upstream, the high integrity valves must be upstream, when it is downstream, the high integrity valves are located downstream of the meters. Provers upstream of turbine meters are not usually recommended, unless the upstream flow conditions when proving and not proving are identical. Other block valves are desirable for isolation of the meter run for maintenance, however these are not required to be of high integrity design as they are normally left in the open position. If a prover is shared with multiple outlet headers a block and bleed valve is required between the prover outlet and each outlet header.

303

Control Valves When multi-meter runs are used, flow control valves are usually required to balance flow among the parallel runs and control flow for proving. Due to the prover pressure drop being added to the meter run pressure drop when proving, the control valves in the meter runs not being proved must be pinched down to force the desired flow rate through the prover and meter being proved. In addition, if turbine meters are being used, downstream back pressure must be kept on the turbine meters for proper operation.

Figure 3 Typical Meter / Prover System P&ID Minimum back pressure is higher for liquids with higher vapor pressure. In order to accomplish the above, a combination back pressure and flow control system is commonly used. It is important to note that the control valves mentioned above may also be used to drop excessive upstream pressure and control on and off for batching. If this added function is required of these valves, it must be known before they are selected, as valves with these added functions are normally considerably larger and more costly than those selected for merely balancing. Consideration must also be made to failure mode on loss of air or signal for the control valves. Usually, failure in last position is most desirable for metering applications.

304

Strainers Meters, provers and the integrated in-line components must be protected from harmful debris and particulate matter that may come from the inlet lines in the form of pipe rouge, loosened internal lining from pipe or storage tanks, "grass", remains from construction, slag, weld rods, etc. This is normally done using large, basket-type strainers. Each is usually equipped with differential pressure switches or transmitters and/or indicators to alarm for a clogged condition. Transmitters are preferred for this application as then computer readout systems can show trends for this important operational parameter. For use with turbine meter systems, special high capacity, low pressure drop strainer designs are available. Skid Instrumentation Meter skids can be operated locally using manual and/or motor operated valves, meter mounted registers, local prover counters, pressure indicators and temperature indicators. However, most skids are equipped with the above mentioned items as back up to a remote control panel. Meter skids are normally equipped with temperature and pressure transmitters per meter run and on the prover outlet. If the prover is some distance from the meters, prover inlet instruments are also desirable. Quality Measurement Instrumentation Sampling Meter systems are usually equipped with quality measurement equipment in addition to the meters and prover used for quantitative measurement. Often these are provided as a separate skid and they may be supplied in a building for locations with extreme climates or required by specification. Quality instrumentation can include densitometers, BS&W monitors, viscometers and samplers. Often slip-stream pumps are included in the quality loop to assure constant flow rate through the instruments, independent of header flow rate. If the fluid is being injected into a pipeline based on certain quality criteria, the quality instrumentation is located upstream of the meters and a reject valve arrangement is used. If the quality package is for monitoring only, as is the case for ship loading and unloading applications, the quality skid is normally downstream of the meters. With this arrangement, the mixing action of the meters and intrusive components can be used to assure a homogeneous sample.

Figure 4 QMI diagram

305

Control Room Instrumentation The trend today is toward more sophisticated control room equipment and an increased amount of remotely operated and monitored equipment. With conventional skid instrumentation, the control and signal cables between the skid(s) and the control room become a major initial cost and an ongoing maintenance cost consideration. Skid mounted serial communication equipment for transmitting data generated and control commands are now available. This equipment reduces total installed cost of the system without sacrificing reliability or accuracy. When evaluating system proposals, total installed cost including interconnection cables must be considered. Two operationally diverse systems are usually available: One is for batch type loading or unloading of vessels, barges, tank trucks or rail cars where a target volume is preset and the system starts up each meter run in turn, brings all runs on line, proves them, delivers the prescribed quantity, and shuts itself off with "ramped down" flow rate control followed by printout of complete reports. The other type of system is strictly a recorder of events and is primarily used for pipelines where flow rate, start, and stop are controlled elsewhere.

Figure 5 Control System Block diagram Periodic reports are logged out and meters are proved when predetermined parameter changes occur. It is necessary for the specification resource and for the design engineer to choose one of these two control modes for each specific system. Valve control and interlock both on and off the meter skid are often required as a part of the control system. Control functions may be done with relays but is now more often done using compact programmable logic controllers. A typical system block diagram is shown in Figure 5 and a typical control panel is shown in Figure 6.

306

Todays personal computer based control systems offer such advanced features as trending of meter factors, production of meter control data, archiving of data, and sophisticated graphical man machine interfaces. They also allow for export of databases to upper level DCS, SCADA or otherwise enterprise management computers. Flow computations normally follow API Manual of Petroleum Measurement Standards Chapter 12 and 21 guidelines: however various petroleum companies and government weights and measures authorities have their own special provisions and or methods of doing certain spreadsheet programs for further analysis and enterprise management purposes. The newest API tables are not always universally used and many "new" installations still are contractually required to do calculations using older tables. These requirements must be known and defined prior to finalizing system software.

Figure 6 Typical Control Panel System Data Concentration and Transmission Many measurement systems are tied via Sat Com, microwave, or telephone lines to large accounting and/or master control computers. If this is required the measurement flow supervisory computer then must also act as a data concentrator and communication device to the larger "host" computer. The protocol required for these two machines to talk (hardware, software, and source of power) must be carefully defined so the systems can properly communicate and interact. Conclusion Liquid measurement station design involves a multi disciplined engineering approach in the proper selection of a large number of interacting components to achieve function, accuracy, dependability, ease of operation, reliability, safety, conformance to prescribed standards, and value. Selection of a competent measurement systems vendor experienced in all aspects of such systems is essential to achieve the end result on schedule and within budget. The vendor must be provided with complete information regarding system fluids and performance requirements prior to starting design in order to efficiently and effectively achieve the end result.

307

Acknowledgement The author acknowledges with thanks the permission to publish this paper granted by FMC Technologies Measurement Solutions. This paper was first presented in May 2005 at the 80th International School of Hydrocarbon Measurement in Oklahoma City, Oklahoma by Peter P. Jakubenas. References 1. "Improving Petroleum Custody Transfer Measurement Accuracy", P.D. Baker - Smith Technical Paper 111. 2. "Flow Metering Devices" Description & Features - Smith Technical Paper 110. 3. Fundamentals of Liquid Turbine Meters, Joshua W. Rose - Proceedings of the Seventy Ninth International School of Hydrocarbon Measurement - 2004 p.p. 204 to 210 4. Displacement Meters for Liquid Measurement, James Henderson - Proceedings of the Seventy Eighth International School of Hydrocarbon Measurement - 2003 p.p. 174 to 180 5. Mass Meters for Liquid Measurement, Steven J. Smith - Proceedings of the Seventy Fourth International School of Hydrocarbon Measurement - 1999 p.p. 289 to 294 6. Liquid Flow Provers (Conventional), Richard Keating - Proceedings of the Seventy Seventh International School of Hydrocarbon Measurement - 2002 p.p. 438 to 446

308