Types of Fluid Flow Measurement
Fluid flow measurement is divided into several types, since each type requires specific consideration of such factors as accuracy requirements, cost considerations, and use of the flow information to obtain the required end results.
When deciding on the best type of meter to measure a given flow, the nature of the fluid to be measured needs to be considered. Flow characteristics are also important. In custody transfer metering, the best flow measurement is required, so that the two parties to the transactions are treated fairly. This chapter considers the factors that need to be taken into account when deciding on a meter to measure fluids in different situations.
Keywords
fluid flow; measurement; flow meter; measurement contract; measurement station; accuracy; uncertainty; maintenance; proving
Fluid flow measurement is divided into several types, since each type requires specific consideration of such factors as accuracy requirements, cost considerations, and use of the flow information to obtain the required end results.
What type of flow meter is best? A common question asked at flow measurement schools and seminars is: “What type of meter is best for my application?”
The answer obviously depends on many factors, but the first consideration, which is often ignored, should be about the fundamental nature of the fluid (liquid, gas, or steam) to be measured. Is there flashing or condensing? Are there well-defined pressure, volume, and temperature (PVT) relationships? Does a predictable flow pattern exist based on the Reynolds number (Re)? Is the flow Newtonian? Is the fluid free of foreign materials that will affect meter performance (solids or gas in liquids, or liquids or solids in gas)? Does the fluid have a constant composition or slowly changing measurable analysis?
Flow characteristics are also important. Is there a fairly constant rate, or do rate variations fall within the meter’s response time and measurement range? How about a laminar, irregular, swirling flow pattern? Single or multiphase at the meter inlet? Smooth or pulsating? Will the liquid flow fill the pipe?
Certain types of meters may have special characteristics to handle some of these problems, but extra care should be exercised in evaluating such equipment to ensure successful measurement. It is worth noting that most gases and many liquids that are routinely handled in the field are not as clean as often assumed. Time and again, measurement and meter evaluation projects have been seriously flawed by debris and crud buildup on metering components. An astute measurement specialist never assumes that the line is clean.
It is also important to consider the fluid’s critical temperature and critical pressure. A meter’s specified accuracy is invalid if the fluid to be measured exhibits large volume changes over minor temperature and pressure changes, which is the case near critical conditions.
At a meter station that measures product worth $1 million a day, an inaccuracy of ±0.2% represents $2,000 a day, or $730,000 a year—an amount that justifies considerable investment to improve flow measurement. The same error for a station measuring $1,000 worth of product a day represents only $2 a day, and the law of diminishing returns limits the investment that is justifiable to improve the measurement accuracy.
Custody Transfer
When money is to be exchanged the best flow measurement becomes important, so that the two parties to the transaction are treated fairly. The desired accuracy limit for flow measurement is 100% correct. However, no measurement is absolutely accurate; it is simply accurate to some limit.
In custody transfer metering, the constant awareness that flow measurement equates to dollars changes the perspective accordingly. The goal sought is that custody transfer measurement be converted to dollars ± zero.
Quantities for custody transfer are treated as absolute when they are billed. The responsibility for this measurement, then, is to reduce all inaccuracies to a minimum so that a measured quantity can be agreed upon for exchanging custody.
The accuracy of set point control measurement may be accepted at ±2%; operational measurement may require no more than ±5%, as contrasted with the ±0% target for custody transfer metering (Figure 3-1).
Measurement Contract Requirements
As stated, measurement becomes more complex when two parties must agree to the quantity of product exchanged and agree to pay/receive money based upon this quantity. To protect each party’s interest, a contract is normally written that specifies all requirements for measurement of the delivered material, such as:
All of these items of interest should be settled prior to commencing measurement for custody transfer purposes. A number of these items are typical in any contract, but it is of value to expand on the ones that impact the measurement equipment and procedures.
Quantity of Material
This will specify not only the quantity of the material to be measured by the seller, but also any rights the seller may have to quantities of material above or below the agreed-upon amount. This requires that the responsible measurement personnel are aware of these values, see that contract limits are being met, and ensure that the seller has the capability of meeting them.
Point of Delivery
The contract sets out the point of custody transfer. If the measurement point and the point of delivery are not the same, an agreement must be reached between the buyer and the seller for responsibilities for the material between these two points.
Material Properties
Limits are specified for certain basic properties, such as composition limits, pressure, and temperature, and the actions to be taken if the material is outside the limits.
Measurement Station Design
The ownership and responsibility of both the buyer and seller for the design, installation, maintenance, and operation of the meter station are spelled out. For metering stations covered by standards, specific references to the standards are made. These standards may be government requirements, industry practices, or individual company guidelines; usually they are combinations of these, and detail the kind of meters to be used plus related correcting and readout systems. Details of access by both parties to the equipment and the requirements for frequency of testing and/or reports are spelled out. For large dollar-exchange quantities, there may be an allowance for a check station, with the same provisions listed as above; stating how any discrepancies between the two measurements will be handled (Figure 3-2).
Some provision is made for estimating deliveries during times when the meter is out of service or registering inaccurately, and the procedure for resolving quantities during these periods will be included. Some accuracy limits are set; if these are not met as determined by test or check meters, settlement provisions are implemented. This accuracy limit may typically be from ±0.5 to ±2.0%, but may be set closer or wider depending on the economics and measurement ability of the meters.
A time period during which a correction can be made is stated if it is not possible to determine the error source and the time of change. This is normally restricted to one half of the time since the last calibration. Requirements for retaining records and reports are spelled out for both parties. This relates to the specified time period allowed for the quantity measurement to be questioned, if major errors are found covering volumes over many months or even years.
Measurements
This specifies, in non-confusing terms, the unit of quantity to be delivered. In a volume measurement, base conditions of temperature and pressure are clearly defined. In a weight (mass) measurement, only the unit of weight need be specified. For most commercial purposes, the terms “weight” and “mass” are used interchangeably without concern about the effects of the attraction of gravity on the weight being measured. Requirements are specified for all related equipment (beyond the basic meter) and how these secondary measurements will be used to correct basic meter readings. These requirements are particularly important, since interested government parties and the parties to the contract in their plant quantity reports may record data on a different base calculation. Major confusion can arise if all of these requirements are not spelled out and clearly understood so that volumes are given on the same basis.
Material Quality
Any natural or manufactured product can have small and varying amounts of foreign material that are not desirable, or at least whose quantities must be limited. The quality section defines the rights of the buyer and the seller if such limits are exceeded. These specifications may also include separate pricing for mixed streams, so quantities must be delineated for proper payment. If there are too many unwanted contaminants, a price reduction may be allowed rather than curtailing a delivery. These details are spelled out in the quality requirement section.
Billing
This section sets a deadline for the computation of quantity, with a provision for correcting errors. It specifies the procedure for billing, the payment period, and penalties for late payment.
Summary of Contract Requirements
A properly written contract, which protects the interests of the buyer and the seller to allow fair and equitable billings to be made for an exchange of the quantity of material, is a basic requirement for establishing custody transfer. The ultimate definition of measurement accuracy is achieved when the seller sends a bill, the buyer pays the bill, and both parties are satisfied with the results. All possible misunderstandings and means of their solution should be defined in the contract in case there is a disagreement.
Other Factors in Custody Transfer
Accuracy
A term used frequently in flow measurement is “accuracy.” Accuracy is a term more often abused than correctly used. Unfortunately, it is a sales tool used commercially by both suppliers and users of metering equipment. The supplier with the “best” number wins the bid. Likewise, the user will sometimes require accuracies beyond the capabilities of any meter available. In either case, the accuracy definition serves a purpose for the practitioner or supplier, but has little relevance otherwise. In custody transfer measurement, accuracy is usually defined as the difference between the measured value and the true (reference) value expressed as a percentage. The problem with this definition is that the indicated value is read from the meter, but the method of obtaining the true (reference) value may not be specified; therefore, the true (reference) value is not precisely known. For this reason, it is the subject of many disagreements.
The related term of the “uncertainty” in a specified procedure is a statistical statement with at least a comparative meaning when examining various meter capabilities.
Uncertainty
The performance of the measurement under flowing conditions can be evaluated by making an uncertainty calculation. Many calculation procedures are available in the standards and flow measurement literature. One is “Measurement Uncertainty for Fluid Flow in Closed Conduits” ANSI/ASME MFC-2M. Another is the American Gas Association (AGA) Report No. 3, “Orifice Metering of Natural Gas, Part 1, General Equations and Uncertainty Guidelines.” The value of this calculation is not a numerically “absolute” or explicit value (Figure 3-3).
These calculations must consider the particular operating conditions for the specific meter application in order to be most useful in getting the most accurate measurement.
Calculation of the equation’s variables is not the whole concern for complete uncertainty determination; allowance must be made for errors in human interpretation, recorders or computers, installation, and also for fluid characteristics. However, most of these are assumed to be minimized, provided that industry standard requirements are met and properly trained personnel are responsible for operating and maintaining the station. Since these effects cannot be quantified, they are minimized by recognizing their potential existence and properly controlling meter station design, operation, and maintenance. Without proper attention to the total problem, a simple calculation of the variables in the equation may mislead a practitioner into believing measurement is better or worse than it actually is.
The uncertainty calculation assumes that the meter has been properly installed, operated, and maintained. If maintenance is neglected and the meter has deposits on it that change its flow characteristics, then the calculation is meaningless (Figure 3-4).
Maintenance of Meter Equipment
Both the supplier and the customer must have confidence that a billing meter is reading the proper delivery volumes. Equipment calibration may change over a period of time, so both parties should take an active part in periodically testing the meter system. Without tests to reconfirm original accuracies of the metering system, any statement of accuracy is not complete.
Proving Meters
If there is a desire to reduce measurement tolerances, then an actual throughput test can be run against a “master meter” or a prover system. The master meter should be calibrated and certified to some accuracy limit by a testing facility, a government agency, a private laboratory, a manufacturer, or the practitioner, using agreed-upon flow standards. Periodically, the master meter has to be sent back to the testing facility for recertification. The retesting frequency depends on the fluids being tested and the treatment of the master meter between tests. The best throughput test is one that can be run directly in series with a “prover.”
The prover can come in many forms, but essentially it involves a basic volume that has been certified by a government or industry group. Since it is one step closer to a basic calibration, this is the most accurate test of a meter’s throughput. Such provers for liquid may be calibrated seraphin cans (for fluids with no vapor pressure at flowing temperature), pressurized volume tanks (for fluids with vapor pressure at the flow temperature), or pipe provers (formerly called mechanical displacement provers as described in the API Manual of Petroleum Measurement Standards). Such pipe provers may be permanently installed in large-dollar-volume meter stations, or may be portable units for smaller multiple meter stations (Figures 3-5, 3-6).
Proper maintenance and calibration of a billing meter is essential for accurate custody transfer metering. Testing requires that both the supplier and the customer participate. Diagnostics and evaluation with proper test equipment ensure that recorded volumes are correct. Any proving must be documented and signed by both parties, so that contract provisions can be implemented for any corrections required.
Properly trained personnel, who understand the importance of the equipment they maintain, are the key to accurate measurement. With proper test procedures, accurate test equipment, a good maintenance procedure, and a timely test frequency, any company should have an acceptable “lost or unaccounted for” record.
See Chapter 9 on maintenance for further discussion of this vital phase of instrumentation.
Operation Considerations
Recognition of a meter’s operating limits must be considered for a meter to meet its stated accuracy. Most meters operate within stated limits and should not be used in the extremes of ranges for custody transfer metering.
Flow variations should be minimized by better control of the flow rate. If there is no single meter that has the range required to operate in the accurate part of its range, then the use of multiple meters, along with some type of flow switching control to turn meters in and out of service, is required.
In addition to basic meter problems at meter extremes, the secondary equipment that measures pressure, temperature, differential pressure (for head meters), density or specific gravity, and composition of the flow can also have problems. Normal specifications of these devices are stated as percentage of full scale. Selecting an instrument with the wrong range for the parameters to be measured introduces errors (Figure 3-7).
A properly chosen, designed, and installed system may still fail to meet expectations if the meter is not operated in its most accurate range.
Custody Transfer Auditing
When money is exchanged for measured material, the two parties’ agreement will include a means of auditing the volumes obtained. Sufficient operation and maintenance records made available to both parties will ensure that the calculated volumes can be arrived at independently. At least a check of the values used by the other party should be made to ensure that agreement is reached over the volume.
This procedure is an important aspect of custody transfer metering, and is usually completed within 30 to 60 days after a bill is submitted. It will keep both parties involved in the measurement, and will prevent disagreements about procedures and volumes at some later date. With the data still current, a disagreement can be settled while knowledge of the measurement is fresh in both parties’ minds. A complete file of any disagreements must be kept, including resolutions. Records can be reviewed to see whether a particular station or particular errors are involved in recurring problems that need to be addressed by an equipment or maintenance upgrade (Figure 3-8).
Summary of Custody Transfer
Custody transfer measurement begins with a contract between two parties that specifies the data needed to choose a metering system. To get the most accurate (minimum uncertainty) measurement required to minimize settlement problems, maintenance and operation of the system must be controlled so that the accuracy capabilities of the meter may be realized in service. The information in this chapter should be supplemented by reference to the other chapters of the book for a complete understanding of an individual meter’s advantages and limitations. If all precautions are taken, then proof of the station will be when bills are submitted and paid, and the custody has been successfully transferred.
Non-Custody Transfer Measurement
Set Point Control Signal
Of prime importance in any process is the ability to measure flows so that the process can be controlled. The absolute accuracy of such a signal is not as important as the ability to repeat the measurement under the same flow conditions. It is confusing to speak of not needing accuracy for flow signals as much as needing repeatability, but seldom does a process operate at its original design parameters. The process must be tuned before it comes into balance. Then the function of the set point control signal is to hold the balance and make the changes required when the process is varied.
An unpredictable meter output can cause control problems, so a control signal must come from a repeatable measurement at given rates across the range of flow. Because of the sensitivity of some processes to changes, the response time of a flow control signal is much more critical than for a custody transfer. In set point control measurement, the rate signal represents the process variable of interest, whereas in custody transfer, the total flow is required. However, in contrast to custody transfer measurement where the ability to integrate volume over time is a significant part of the system, the set point control signal is seldom integrated to totalize the volume.
Other Uses
Flow may be measured periodically to check an operation with the assumption that it will then run properly until results indicate otherwise. A good example is a heating and cooling distribution system using ducts. Once set, the system is changed only if there is an indication that the distribution has changed as heating or cooling gets out of control. Similar tests are done for pollution studies. In these “other uses,” accuracy may be no better than ±5 to ±10%.
Flow may also be used as an alarm signal to indicate that the design limits of the flow volumes (either high or low) have been exceeded, and action should be taken.
Summary of Flow Measurement
Many different capabilities are required to measure flow. Each application should be defined so that expectations of accuracy can be balanced against cost to derive the most cost-effective installation that will do what is required.