TURBINE FLOW METER FOR LIQUID MEASUREMENT 1. Introduction Turbine flow meters are designed to accurately measure the flow of liquids and gases in pipes. They are volumetric flow measuring devices and have been commercially available since the late 1940s. Sizes exist from a variety of manufacturers to cover the flow range from 0.001 gpm to over-25,000 gpm for liquid service and 0.001 acfm to over 25,000 acfm for gas service. End connections are available to meet the various piping systems. The flow meters are typically manufactured from austenitic stainless steel but are also available in a variety of materials, including plastic. Turbine meters are applicable to all dean fluids over a pressure range from sub-atmospheric to over sixty thousand psi and temperature ranges from cryogenic to about 800°C (1500°F).The turbine flow meter is perf1aps the most accurate type of meter available under specified operating conditions. It is capable of repeating to 0.025% of reading with accuracy and traceability to 0.05% of reading for liquid service; it is also capable of repeating to 0.1% of reading with accuracy and traceability to 0.3% of reading for gas service. A turbine meter has only one moving part - the rotor. Components can be selected for compatibility with most fluids, such as corrosive chemicals, dairy products, cryogenic liquids, steam, fuels, and water (including de-ionized water). The output signal from the flow meter and sensors is an electronic pulse, but other output signals such as analog, visual, or digital are also available. 2. Working principle and design A turbine flow meter consists of a rotor mounted on a bearing and shaft in a housing . The fluid to be measured is passed through the housing, causing the rotor to spin with a rotational speed proportional to the velocity of the flowing fluid within the meter. A device to measure the speed of the rotor is employed to make the actual flow measurement. The lack of a velocity component in the tangential direction is an important element in the system and will be discussed in more detail. The rotor blade is oriented at an incidence angle to the flow stream. The momentum of the flowing fluid imparts a rotational velocity on the rotor, causing the flow to change direction and depart from the rotor in a swirling direction. The rotational velocity is nearly directly proportional to the velocity or flow rate While the blade in the turbine meter works much like a typical turbine blade, there are some significant differences. A standard turbine blade is designed to do work by turning a shaft from which a major portion of the energy in the system is extracted. It is therefore necessary to design the most efficient blade possible. The rotor of a conventional power producing turbine is usually designed to operate at a specific speed or over a limited range of speeds. Seldom is a power producing turbine allowed to operate at its "free-spin speed".

The rotor in a turbine meter is not attached to a shaft: to produce useful work and is always operating at its "free-spin speed". The range in speed over which the turbine flow meter rotor operates is usually quite wide. The most important design characteristic of the turbine rotor is stability, not efficiency as with a drive turbine. A good turbine flow meter must transform an axial velocity in to a rotational velocity as precisely and repeatable as possible. It is not necessary to produce the highest free-spin speed possible, but to produce the most stable representation of the axial velocity and flow that is the least affected by secondary influences. Again, for a turbine meter, the most important characteristic is stability, and the design parameters that make an efficient blade also make the blade sensitive to minor changes in environmental conditions such as velocity profile. Another desirable characteristic of a turbine meter is to maintain the meter constant (K factor) at the same value over as wide range of flow as possible. A highly efficient blade tends to vary in efficiency as a function of speed (flow rate), resulting in a significant decrease in meter constant as a function of flow rate. In reality, the inefficient blade designs are better for a turbine meter. A blade design that does not change its efficiency with operating conditions, such as velocity, velocity profile , temperature, or density , is best. In other words, a design that is minimally sensitive to changes in Reynolds number is optimum. While the design details can be tailored to minimize the effects of Reynolds number on a turbine meter, it is more difficult to minimize the effects of a swirling component to flow. Since a turbine meter contains a rotating component, it is sensitive to a swirling component of the flow velocity. The upstream support of the meter is used by most manufacturers to "help" reduce any swirl, but it cannot be used to completely eliminate swirl dueto flow meter size limitations. The drag imposed on the rotor by the other components in the, meter greatly affect the operation of the flow meter, especially at the lower end of its operating limits. The bearings are one of the largest contributors to drag. If the meter has a mechanical type pick off it can impart drag on the rotor as well. The larger the meter, however, the less sensitive it is to these drag forces because there is proportionally more blade area, generating more torque to drive the rotor. However, in small meters, especially in gas applications, bearing and the sensor drag is of prime importance at lower flow rates. The hangers or support vanes that position the bearings centrally in the pipe are commonly used as flow strengtheners to reduce swirl. Their shape varies considerably. In fig.1, they are flat plates, and in another design, they are parallel tubes. The rotation of the rotor is sensed by a pickup, which most commonly senses a change in the magnetic material permeability in its vicinity as a blade passes (magnetic reluctance change). Alternatives to this are to use the eddy current effect or dynamo effect as a blade passes through the magnetic field of the sensor (magnetic inductive) or to use a modulated high frequency signal (modulated carrier or radio frequency). The capacitance effect can also

be used to sense the proximity of a passing blade to the sensor head. Optical fibers have been tried, but if the optical path passes through the fluid, there is a danger of the windows becoming fouled. For increased integrity, some meters have two sensors spaced at about 120o, the pulse trains from both are sensed, and the phase angle is constantly checked. In small meters, the internals are held in position by circlips, the removal of which allows the internals to be withdrawn. In other designs and larger sizes, the bearing assemblies may be removed from the body by releasing a taper locking nut. 2. Elements of construction The turbine flow meter consists of a housing or body, a rotor supported by bearing(s), a shaft, an upstream support structure, usually a downstream support structure, a sensing device, and a readout device(s). Some meters have two rotors or other unique features. As with any flow meter, the design of a turbine flow meter is a compilation of compromises that result in the best overall design in terms of performance and cost. Each designer may weigh the various factors differently in terms of goals and experience. There is no "correct" design. All will contain certain advantages and work well within certain limitations, given an associated level of cost. It is the understanding of these advantages and limitations and the relative cost of each feature that make the selection of the best meter system possible. 2.1Bearing Design Materials Many researchers have attributed calibration unpredictability to bearing variation as a major factor. Clean Liquids Open ball bearings of stainless steel or other compatible material are suitable for liquids with lubricating properties in the temperature range -50 to 250oC, but a filter with a mesh size small enough to prevent solid particle ingress into the bearings should be used. Liquids in General Tungsten carbide or high chrome / high cobalt journal bearings possibly with carbide pinions and a satellite sleeve are suitable in the range -50 to 400oC unlubricated. The bearings are lubricated by the metered fluid and made up of tungsten carbide with a hardness of Rockwell C-94 and a surface smoothness of 0.05μm. Tungsten carbide with cobalt bonding is most common, but the cobalt is leached out by acidic solutions used for cleaning. Stellite may then be used as a less durable alternative for more corrosive fluids. A more recent alternative is nickel-bonded carbide bushes with titanium carbide shafts. An alternative design uses rotor pins of hard metal in sapphire bearings with thrust plate and rings made with curved contact surfaces for line contact. Ploytetrafluroethylene

(PTFE) is used in some applications. Ceramic bearings (e.g., Al2O3) may also be an option. One manufacturer has used a ball and sleeve design. The shaft was tipped with an ellipsoid and ran in a ceramic sleeve so that line contact was achieved. Typical bearing life was claimed as 12-24 months when used with 50-75% duty cycle in clean particulate-free gasoline. For ball race bearings correctly lubricated, a life of 4,000-6,000 hours was claimed, and for journal bearings 20,000 hours or more was claimed, depending upon the properties of the measured fluid and rate of flow. For some larger sizes (above about 80mm) only an upstream bearing is used; it consists of a tungsten carbide (or other suitable materials) bush and shaft. Special bearings are available for cryogenic applications. Hydrostatic bearings may be suitable for dirty fluids. House and Johnson (1986) applied hydrostatic bearings and obtained a remarkable performance (turndown ratio for water of 1,000:1) and suitability for dirty fluids. Some manufacturers provide application charts. Ball bearings have been found satisfactory for cryogenic applications in rocket testing because of their ability to tolerate over speed and operation with gas. However, journal bearings, although less linear, may be more repeatable. 3.2 Strainers Typical values for strainers are given by manufacturers with maximum allowable particulate sizes (Table 1). Table1 - Typical Values for Strainers from Manufacturer's Literature with Maximum allowable particulate sizes Meter Diameter (mm)

Particulate Sizes (μm) Mesh

Hole Size (mm)

Ball Bearings

Journal Bearings

13 60 100 150 0.100 20 90 142 100 0.150 25-80 100 185 80 0.175 100-150 142 251 60 0.200 200-300 251 401 40 0.300 3.3 Size Ranges Meter sizes are typically: for flanged, 6 - 500mm for threaded, 6 - 50mm In some designs, the bore reduces so that at the propeller the bore radius is a minimum. Flow ranges are typically 10:1 turndown (but may be up to 30:1) for flow rates of 0.03 -

7,000 m3/h. Over speed up to 1.5 times the maximum flow rate may be permitted for short periods, but if it takes place for prolonged periods, the bearing life is reduced. Pressure losses at maximum flow rate, with water as the flowing medium, range up to 0.2 bar ( 2 x 104 N/m2) in 12mm sizes and up to 0.25 bar (2.5 x 104 N/m2) in 200mm sizes. Maximum pressures ranges up to 25 bar for hygienic designs, up to 240 - 400 bar for threaded designs, and according to the flange rating for flange designs. Temperature ranges are typically -50 to 150oC but may be as wide as -265 to +310oC. Some manufacturers give plots of the effect of change of viscosity on the performance of their meters. The bearing, as well as the blade design, will have an influence on this. Acceptable viscosity limits for smaller sizes (less than 80mm) may be up to 15 mPas (15 cP) and for sizes above 100mm up to 50 mPas (50 cP). These viscosity limits prevail as a result of increasing non linearity with reducing Reynolds Number. 3.4 Other Mechanical Design Features Internals may be made of stainless steel, but other materials such as Teflon and Polythene are also used. In order to increase the pulse rate from impellers with typically six to eight blades, a shroud ring with holes or high permeability soft iron alloy studs is used. This increases the number of pulses due to the holes changing the magnetic reluctance and also allows viscosity compensation to improve low speed performance. The shroud can also give greater strength and reduced blade vibration. In some small designs of turbine meter, the blades have a T shape so that at the tip they are longer (longer chord) than at the hub. One reason given for this is to improve insensitivity to viscosity change, which may result from a greater driving torque achieved by increasing the blade lift at the maximum radius. It may also flatten the characteristic hump at low speed. Some designs can be used bidirectional, but, of course, they require calibration in each direction. Response time constants measured with water or with a liquid of similar density range from 0.005 to 0.05 s for a 50% flow rate change or up to 0.17 s to reach 63% of a step change final value. After use with corrosive fluids, the meter should be cleaned with solvents. However, maximum rotational speeds should be carefully observed. 4.Design and operational considerations in meter turbine flow installations The design of turbine meter installations should take into account the following considerations .The installation should be able to handle the maximum and minimum flow rates , the maximum operating pressure ,and the temperature range and the type of liquid to be measured .If necessary the installation should include protective devices that keep the operation of the meter within design limits. The installation should ensure a maximum dependable operating life . Strainers or air eliminators may be provided

upstream of the meter to remove solids that could cause premature wear or gases that could cause measurement error .The installation should ensure adequate pressure on the liquid in the metering system at all temperatures so that the fluid being measured will be in liquid state at all times. The installation should provide for proving each meter and should be capable of duplicating normal operating conditions at the time of proving. The installation should ensure appropriate flow conditioning both upstream and downstream of the meter or meters. The installations should comply with all applicable regulations and codes. 4.1 Installation Flow conditioning: The performance of turbine meters is affected by liquid swirl and non uniform velocity profiles that are induced by upstream and downstream piping configurations valves, pumps, joint misalignment, protruding gaskets, welding projections, or other obstructions. Flow conditioning shall be used for swirl and non uniform velocity profiles. Flow conditioning requires the use of sufficient lengths of straight pipe or a combination of straight pipe and straightening elements that are inserted in the meter run upstream(and sometimes downstream) of the turbine meter When only straight pipe is used , the liquid shear or internal friction between the liquid and the pipe wall shall be sufficient to accomplish the required flow conditioning .Experience has shown that in many cases 20 D upstream and 5 D downstream of the meter provide effective conditioning. A straightening element usually consists of a cluster of tubes, vanes or equivalent devices that are inserted longitudinally in a section of straight pipe . Straightening elements effectively assist flow conditioning by eliminating liquid swirl. 4.2 Piping installation Turbine meters are normally installed in a horizontal position. The manufacturer shall be consulted if space limitations dictate a different position. Where the flow range is too great for anyone meter or its prover, a bank of meters may be installed in parallel. Each meter in the bank shall operate within minimum and maximum flow rates. A means shall be provided to balance flow through each meter. Meters shall be installed so that they will not be subjected to undue stress, strain, or vibration. Provision shall be made to minimize meter distortion caused by piping expansion and contraction. Measurement systems shall be installed so that they will have a maximum, dependable operating life. This requires that in certain services protective devices be installed to remove from the liquid abrasives or other entrained particles that could stop the metering mechanism or cause premature wear. If strainers, filters, sediment traps, settling tanks, water separators, a combination of these items, or any other suitable devices are required, they shall be sized and installed to prevent flash vaporization of the liquid before it passes through the meter. Protective devices may be installed singly or in an

interchangeable battery, depending on the importance of continuous service. In services where the liquid is clean or the installed meter does not require or warrant protection, omission of protective devices may be acceptable. Monitoring devices should be installed to determine when the protective devices needs to be cleaned. Measurement systems shall be installed and operated so that they provide satisfactory performance within the viscosity, pressure, temperature and flow ranges that will be encountered. Meters shall be adequately protected from pressure pulsations and excessive surges and from excessive pressure caused by thermal expansion of the liquid. This kind of protection may require the installation of surge tanks, expansion chambers, pressure-limiting valves, pressure relief valves, and/or other protective devices. When pressure relief valves or pressure-limiting valves are located between the meter and the prover, a means of detecting spills from the valves shall be provided. Conditions that contribute to vaporization of the liquid stream shall be avoided through suitable system design and through operation of the meter within the flow range specified by the manufacturer. Vaporization can be minimized or eliminated by maintaining sufficient back pressure in and immediately downstream of the meter. This is generally accomplished by placing a back-pressure valve downstream of the meter to maintain pressure on the meter and the prover above the vapor pressure of the liquid. In some operations, the normal system pressure may be high enough to prevent vaporization without the use of a back Pressure Valve. For low-vapor-pressure liquids, the numerical value of the minimum back pressure should be calculated as follows: P b = 2∆ p + 1.25 Pv Where , Pv = Absolute vapor pressure at the maximum operating temperature, in psi absolute P b = Minimum back pressure in psi gauge ∆ p = Pressure drop across the meter at the maximum rate of flow in psi With high-vapor-pressure liquids, it may be possible to reduce the coefficient of 1.25 to some other practical and operable margin. In either case, the recommendations of the meter manufacturer should be considered (Figure 2). When a flow limiting device or a restricting orifice is required, it should be installed downstream of the meter run. An alarm may be desirable to signal that flow rates have fallen below the design minimum. If a flow limiting or other pressure reducing device is installed on the inlet side of the meter, it shall be installed as far as possible upstream of the meter run and shall maintain enough pressure on the outlet side of the meter run to prevent any vaporization of the metered liquid .Each meter shall be installed so that neither air nor vapor can pass through it. If necessary , air/vapor elimination equipment shall be installed upstream of the meter. The equipment shall be installed as close to the meter as is consistent with good practice, but it must not be so close that it creates swirl or a distorted velocity profile at the entry to the meter. Any vapors shall be vented in a

safe manner .Meters and piping shall be installed so that accidental drainage or vaporization of liquid is avoided. The piping shall have no unvented high points or pockets where air or vapor could accumulate and be carried through the meter by the added turbulence that results from increased flow rate. The installation shall prevent air from being introduced into the system through leaky valves, piping, glands of pump shafts, separators, connecting lines and so forth. Lines from the meter to the prover shall be installed to minimize the possibility of air or vapor being trapped. Manual bleed valves should be installed at high points so that air can be drawn off before proving. The distance between the meter and its prover shall be minimized. The diameter of the connecting lines shall be large enough to prevent a significant decrease in flow rate during proving. Flow-rate control valves may be required downstream of each meter , particularly in multi meter installations, to keep the proving flow rate equal to the normal operating rate for each meter. Piping shall be designed to prevent the loss or gain of liquid between the meter and the prover during proving .Special consideration should be given to the location of each meter, its accessory equipment, and its piping manifold so that mixing of dissimilar liquids is minimized. Most turbine meters will register flow in both directions, but seldom with identical meter factors. If flow must be restricted to a single direction because of meter design, flow in the opposite direction shall be prevented. A thermometer, or a thermometer well that permits the use of a temperature-measuring device, shall be installed in or near the inlet or outlet of a meter run so that metered stream temperatures can be determined. The device shall not be installed upstream within the flow-conditioning sections or downstream closer than the manufacturer's recommended position. If temperature compensators are used, a suitable means of checking the operation of the compensators is required To determine meter pressure, a gauge, recorder, a transmitter of suitable range and accuracy shall be installed near the inlet or outlet of each meter 4.3 Electrical installations Turbine meters usually include a variety of electrical or electronic accessories, as discussed above. The electrical systems shall be designed and installed to meet the manufacturer's recommendations and the applicable hazardous area classifications and to minimize the possibility of mechanical damage to the components. Since turbine meters usually provide electrical signals at a relatively low power level, care must be taken to avoid signal and noise interference from nearby electrical equipment. 5. Meter performance Meter performance is defined by how well a metering system produces, or can be made to produce, accurate measurements.

5.1 Meter factor Meter factors shall be determined by proving the meter under conditions of rate, viscosity, temperature, density and pressure similar to those that exist during intended operation. Meter performance curves can be developed from a set of proving results. The curve in figure 3 is called a linearity curve. The following conditions may affect the meter factor: • Flow rate • Viscosity of the liquid • Temperature of the liquid • Density of the liquid • Pressure of the flowing liquid • Cleanliness and lubricating qualities of the liquid • Foreign material lodged in the meter or flow-conditioning element. • Changes in mechanical clearances or blade geometry due to wear or damage. • Changes in piping, valves, or valve positions that affect fluid profile or swirl. • Conditions of the prover 5.1.1 Causes of variations in meter factor Many factors can change the performance of a turbine meter. Some factors, such as the entrance of foreign matter into the meter, can be remedied only by eliminating the cause. Other factors, such as the buildup of deposits in the meter, depend on the characteristics of the liquid being measured; these factors must be overcome by properly designing and operating the meter system. The variables that have the greatest effect on the meter factors are flow rate, viscosity, temperature and foreign matter (for example, paraffin in the liquid). If a meter is proved and operated on liquids with inherently identical properties, and operating conditions such as flow rate remain similar, the highest level of accuracy can be anticipated. If there are changes in one or more of the liquid properties or in the operating conditions between the proving and the operating cycles, a change in meter factor may result, and a new meter factor must be determined. 5.1.1.1 Variations in flow rate At the low end of the range of flow rates, the meter factor curve may become less linear and less repeatable than it is at the medium and higher rates (see figure 3, applications A and B). If a plot of meter factor versus flow rate has been developed for a particular liquid and other variables are constant, a meter factor may be selected from the plot for flow rates within the meter's working range; however, for greatest accuracy, the meter should be reproved at the new operating flow rate.

5.1.1.2 Variations in viscosity Turbine meters are sensitive to variations in viscosity. Since the viscosity of liquid hydrocarbons changes with temperature, the response of a turbine meter depends on both viscosity and temperature. High-density liquid hydrocarbons typically present the greatest problem. The viscosity of light hydrocarbons such as gasoline essentially remains the same over wide temperature changes, and the meter factor remains relatively stable. In heavier, more viscous hydrocarbons such as crude oils, the change in meter factor can be significant because of the viscosity changes associated with relatively narrow temperature changes. It is advisable to reprove the meter under conditions that closely approximate normal operating conditions. 5.1.1.3 Variations in temperature In addition to affecting changes in viscosity, significant variations in the temperature of the liquid can also affect meter performance by causing changes in the physical dimensions of the meter and in the apparent volume measured by the meter as a result of thermal expansion or contraction of the liquid. The tables and formulas in chapter 11 may be used to calculate the extent of liquid expansion or contraction. For greatest accuracy, the meter should be proved in the range of normal operating conditions. 5.1.1.4 Variations in density A change in the density of the metered liquid can result a significant differences in meter factor in the lower flow ranges, thereby requiring the meter to be proved. For liquids with a relative density of approximately 0.7 or less, consideration must be given to raising the value of the meter's minimum flow rate to maintain linearity. The amount of increase in lower flow rates will vary depending on meter size and type. To establish the minimum flow rate, several provings should be made at different rates until a meter factor that yields an acceptable linearity and repeatability can be determined. 5.1.1.5 Variations in pressure If the pressure of the liquid when it is metered varies from the pressure that existed during proving, the relative volume of the liquid will change as a result of its compressibility . (The physical dimensions of the meter will also change as a result of the expansion or contraction of its housing under pressure}. The potential for error increases a proportion to the difference between the proving and operating conditions. For greatest accuracy, the meter should be proved at the operating conditions (Please refer chapter 4 and chapter 12 of MPMS}.Volumetric corrections for the pressure effects on liquids with vapor pressures above atmospheric pressure are referenced to the equilibrium vapor

pressure of the liquid at the standard temperature, 60°F ,15°C or 20°C, rather than the atmospheric pressure, which is the typical reference for liquids with measurement-temperature vapor pressures below atmospheric pressure. Both the volume of the liquid in the prover and the registered metered volume are corrected from the measurement pressure to the equivalent volumes at the equilibrium vapor pressure at the standard temperature, 60°F ,15°C or 20°C.This is a two-step calculation that involves correcting both measurement volumes to the equivalent volumes at equilibrium vapor pressure at measurement temperature. The volumes are then corrected to the equivalent volumes at the equilibrium vapor pressure at the standard temperature, 60°F, 15°C or 20°C. A detailed discussion of this calculation is included in chapter12.2 of MPMS. 6.Operation and Maintenance This section covers recommended operating and maintenance practices for turbine meters. All operating data pertaining to measurement, including the meter-factor control charts, should be accessible to interested parties.6.1 Conditions that affect operation. The overall accuracy measurement by turbine meter depends on the condition of the meter and its accessories, the temperature and pressure corrections, the proving system, the frequency of proving, and the variations, if any, between operating and proving conditions. A meter factor obtained for one set of conditions will not necessarily apply to a change set of conditions .Turbine meters should be operated within the specified flow range and operating conditions that produce the desired linearity of registration. They should be operated with the equipment recommended by the manufacturer, and only with liquids whose properties were considered in the design of the installation .If a bidirectional turbine meter is used to measure flow in both directions, meter factors shall be obtained for each direction of flow. The meter factors can be determined by a prover that has proper man folding and the required protective equipment and flow conditioning located both upstream and downstream of the meter . Failure to remove foreign matter upstream of a turbine meter and its flow-conditioning system may result in meter damage or miss measurement . Precautions should be taken to prevent the accumulation of foreign material, such as vegetation, fibrous materials, hydrates and ice, in the turbine-meter run. 7. Precautions for operating newly installed meters When a new meter installation is placed in service, particularly on newly installed lines, foreign matter can be carried to the metering mechanism during the initial passageofliquid. Protection should be provided from malfunction or damage caused by foreign matter, such as slag, debris, welding spatter, thread cuttings and pipe compound. Following are suggested means for protecting the meter from foreign matter • Temporarily replace the meter with a spool.

• Put by a bypass around the meter. • Remove the metering element. • Install a protective device upstream of the meter. 8. Meter Proving • Each turbine-meter installation should contain a permanent prover or connections for

a portable prover or master meter. The selection of proving methods shall be acceptable to all parties involved.

• The optimum frequency of proving depends on so many operating conditions that it is unwise to establish a fixed time or throughput interval for all conditions. In clean liquid service at substantially uniform rates and temperatures, meter factors tend to vary little, necessitating less frequent meter proving. More frequent proving is required with liquids that contain abrasive materials, in LP gas service where meter wear may be significant, or in any service where flow rates and/or viscosities vary substantially . Likewise, frequent changes in the type of product necessitate more frequent provings. In seasons of rapid ambient temperature change, meter factors may vary accordingly and proving should be more frequent. Studying the meter-factor control chart or other historical performance data that include information on liquid temperature and flow rate will aid determination of the optimum frequency of proving. Provings should be frequent (every tender every day)when a meter is initially installed. After frequent proving has shown that meter-factor values for any gas-liquid are being reproduced within narrow limits, the frequency of proving can be reduced if the factors are under control and the overall repeatability of measurement is satisfactory to the parties involved. A meter should always be proved after maintenance. If the maintenance has shifted the meter-factor values, the period of relatively frequent proving should be repeated to set up a new data base by which meter performance can be monitored. When the values have stabilized, the frequency of proving can again be reduced.

9. Meter maintenance For maintenance purposes, a distinction should be made between the parts of the system that can be checked by operating personnel (parts such as pressure gauges and mercury thermometers) and more complex components that may require the services of technical personnel. Turbine meters and associated equipment can normally be expected to perform well or long periods. Indiscriminate adjustment of the more complex parts and disassembly of equipment are neither necessary nor recommended. The manufacturer’s standard maintenance instructions should be followed. Meters stored for a long period shall be kept under cover and shall have protection to minimize corrosion .Establishing a definite schedule for meter maintenance is difficult, in terms of both time and throughput,

because of the many different sizes, services and liquids measured. Scheduling repair or inspection of a turbine meter can best be accomplished by monitoring the meter-factor history for each product or grade of crude oil. Small random changes in meter factor will naturally occur in normal operation, but if the value of these changes exceeds the established deviation limits, the cause of the change should be investigated and any necessary maintenance should be provided. Using deviation limits to determine acceptable normal variations trikes a balance between looking for trouble that does not exist and not looking for trouble that does exist. 10. Turbine meter accuracy- "degradable" Only if properly applied can turbine flow meters provide accurate flow measurement over wide flow ranges. Guidelines for estimating the magnitude of influences that affect performance need to be understood, especially where accuracy is critical. The axial type turbine flow meter is perhaps the most commonly used transducer which generates a frequency proportional to flow rate. It is simple, small in size and high accuracy , with wide range ability , fast response and relatively low pressure drop. Turbine meters consist of a flow housing in which is mounted a bladed turbine supported by ball, l sleeve or jeweled bearings. the turbine rotates at an angular velocity proportional to the volumetric rate of flow. Accuracies of:!: 0.25% and repeat abilities of :1:0.05% can be attained over a 20:1 range but only when all factors which are known to influence the performance of the meter are controlled. Turbine meter manufacturers usually provide calibration factors based on water, which has a viscosity of one centistokes and a specific gravity of 1.0. In addition, during calibration special efforts are made to provide an ideal, steady flow to the meter, free of entrained air and contaminants. line pressures are kept sufficiently high to prevent cavitations at maximum flow rates and water temperature is close to ambient. Also the meter is usually calibrated in the horizontal position. However important factors affecting performance in the plant include changing viscosities and densities, the effects of upstream and downstream piping, entrained gases, fluid contaminants, liquid pressure, cavitations , liquid temperature and special orientation of the meter, to name but a few. 10.1 Viscosity/density alters ranges Fluid viscosity is a critical parameter in applying turbine flow meters. As viscosity increases, the performance at the lower portion of the flow range is affected more than the higher portion of the range. As a rule-of thumb, the maximum viscosity in centistokes which can be used with the meter still maintaining adequate performance is 2.5 times the meter size. For example a four inch meter would maintain close to 0.5% linearity up to a viscosity of 10 centistokes. Although turbine meters do become more non-linear as viscosity increases, they still maintain repeatable performance characteristics. Liquid turbine meters are also influenced in the lower portion of the range by liquid density. For

specific gravities below 0.70, a rough estimate of the reduction on meter range is given by multiplying the published range by the square root of the SG. For example the linearity specification on a one inch turbine meter is given over a range of six to sixty GPM. This is a linear range of ten to one. If, however, a light hydrocarbon with specific gravity of 0.50 is to be metered, the linear range should be reduced to seven to one. It is possible to maintain the 10 to one linear range by supplying a meter with slightly modified design and / or shifted flow range. 10.2 Entrance conditions are critical Flow meters are usually calibrated in flow standards where the entrance flow to the meter has a fully developed turbulent flow profile free of swirl. Such ideal conditions are developed in long straight pipes, but in practice flow meters must often be located close to pumps or pipe fittings, such as reducers, elbows, valves etc. which disturb the entrance flow pattern, thus impairing the flow meter accuracy . Depending on the meter design and the degree of internal straightening provided, most suppliers recommend either 10 or 20 diameters of straight pipe upstream of the meter and five diameters of straight pipe downstream in order to preserve the meter performance. The 10-20 diameter criterion generally results in installation errors which are well within : t 1 %, with the exception of V4 inch open gate valve and two 90° offset bends set-ups. If severe swirl is present, it is necessary to use a flow straightener upstream of the flow meter. For maximum accuracy , the flow straightener and flow meter should be calibrated as a single meter run to avoid straightener induced shifts in the mean calibration factor. Several straigtener types are available, ranging from the simple tube bundle to the more sophisticated types, such as the high pressure drop Zanker and the interesting Mitsubishi flow profiler. 10.3 Air eliminators for two phases Experience with turbine flow meter calibration facilities has indicated that even small amounts of air entrained in the calibrating fluid (water) can cause K-factor readings to be several percent high. This conclusion has been quantitatively verified . The available data indicates that with as little as 2% air (by volume) in water, most flow meters had over registration errors in the neighborhood of one to 10%. Accurate metering requires installation of air eliminators so that air or entrained gas is not present in the metered fluid. Another form of two-phase flow occurs when the line pressure is low enough to induce cavitations. 10.4 Strainers for contaminants Hard abrasive particles in the process stream are the major cause of turbine flow meter bearing wear. Particles may jam the bearings, particularly ball bearings. To minimize

solids contamination, strainers should be installed upstream of the flow meter. Extremely hard (Rockwell A90 to A92) sleeve bearings utilizing tungsten carbide have been successfully used In process streams contaminated with sand, rust and other solid particles. Bearing life in years has been reported in sand-Laden crude oil. Experience has also shown that these bearings can be operated virtually dry in non-lubricating fluids, such as ammonia and light hydrocarbons. Another point to consider is the possible formation of gum, varnish or other deposits on the bearings if the meter is allowed to dry. These deposits can affect low-end performance for a short time commencing with initial reuse of the meter. If such non-filterable contaminants are present in the process stream, the piping containing the flow meter should be filled with liquid at all times. 10.5 Changing dimensions If readout in gravimetric units is desired, then the process liquid temperature must be determined and density calculated, or a densitometer utilized to provide direct density readings. Temperature does affect flow meter K factor through dimensional changes in the rotor and housing. An approximate equation for the effects of thermal expansion is given by ∆K/K=3β*(T - T 0) where ∆K/K is the fractional change in K factor, β is the thermal expansion coefficient of the flow meter housing, T0 is the temperature of the calibrating liquid and T is the temperature of the flowing liquid. 10.6 Orientation changes calibration Best accuracy is attained when the meter orientation in the field piping is the same as the orientation during calibration. The changes in K factor are mainly caused by changes in bearing loads with meter position. These changing loads alter the frictional retarding torques on the rotor. The effects are seen mostly in the lower one third of the flow range, with the magnitude and direction of the error being totally dependent upon the particular design of the flow meter being used. In conclusion, although many limitations of turbine meters have been discussed, it is believed that they are not any greater than those encountered with other types of flow meters. Some other turbine meter characteristics have not been considered. These include, effects of over range, reverse flow and high line pressures; effects of fluid lubricity on small size meters; effects of fluctuating flow conditions; and the effect of acceleration and vibration. All of these are of interest in certain specific applications. Finally the turbine meter remains the meter of choice in many applications, ranging from batching and blending to custody transfer in refined product pipelines - in fact wherever accurate flow measurement is necessary .

11. Performance / Limitations The accuracy of an instrument is an often misunderstood and misused term. It implies how well the reading presented by the meter represents the absolute truth. However often the user does not use the instrument within its limitations and the measurement is not representative of the instrument's true capability . The manufacturer advertises the potential accuracy of the instrument under ideal conditions. While the instrument may have certain Ideal potential, It may be unrealistic to assume that any given user can achieve that potential in the real world. The accuracy of a turbine flow meter in general is a combination of three elements.

• Repeatability • Calibration and • Application effects

Repeatability can be determined by comparing the repeated results of a meter to a standard. Accuracy of a turbine flow meter is not a function of the meter itself; It is a function of the traceability of the standard against which it is calibrated. Application is the third aspect of accuracy of a turbine flow meter .Effects such as improper or no straightening sections .Operation of the meter at Reynolds numbers different from those of calibration. Not compensating for extremes in operating conditions Use of fixed K factor electronics on a meter used in extended ranges can introduce errors in to the meter.

(Kmax -Kave ) % Linearity = -------------------- * 100

2*Kave The linearity of a TFM is typically quoted as ±0.5% of reading for liquid service by most manufacturers over a 10: 1 range at constant viscosity and temperature. When operating over a wider flow range the linearity will increase. Linearity is also dependant on the viscosity of the fluid .Most TFM 's can be linearized at any viscosity up to approximately 30 cst over a 10:1 range but will become nonlinear at other viscosities. The deviation in linearity can change as much as 30 % over a viscosity change from l cst to 100 cst at the lowest operating frequencies but may be as little as only 1% at the highest operating frequencies. Other factors affecting linearity is The type of pick up used The type of bearing used

12. Specifications When purchasing a TFM it is necessary to give as much information as possible to the manufacturer. This is especially important in purchasing an instrument where a large number of options are available. It is not a good idea to order a flow meter from a manufacturer by a part number or model number unless one is very sure what is being ordered. lf the user does the meter selectio the manufacturer will ship what is being ordered, whether it correct for the application or not. It is best to discuss with the manufacturer and jointly select the meter. The more information one gives to the manufacturer the better the opportunity to get the best meter for the application and more important yet, the better the opportunity to get proper calibration. The following information is required to. Select the meter properly Flow rate range: Maximum, Minimum and Nominal if applicable Temperature: Maximum and Minimum fluid temperature and range of ambient temperature. Pressure: Maximum and Minimum pressure in the fluid stream Type of service : The type of fluid media by descriptive name ,specification , etc

• .End fitting type • Pick off type • Power source

Other information : Explosion proofing ,intrinsically safe ,etc. Pressure drop Materials of construction Fluid properties: Specify the absolute viscosity and density I the kinematic viscosity of the fluid as well as the range of this parameter Units of flow required Electronic output type Electronic enclosure type PERFORMANCE SPECIFICATIONS Linearity Repeatability Accuracy

Linearity B

Application B

Flow rate or Reynolds number

Fig.2 Turbine-Meter Performance Characteristics

Flow range at designated linearity

Flow range at designated linearity, Application A

Linearity A

Increase

Decrease

Meter

factor

unit volume

Pulse per

Increase

Decrease

Manufacturer's statedmaximum flow rate

Back pressuretoo low

Curve representsCavitation

Back pressureadequate

Pulsesper

unit volume

Flow rate or volume per unit of timeNote: All curves are for example only.

Fig.3 Effects of Cavitation on Rotor Speed

Annexure-I Factors Affecting Proper Installation Flow Conditioning/Flow Straightening Under normal conditions a straight run of 10 diameters upstream and 5 diameters downstream is sufficient. The upstream section should have a straightening element. Under extreme conditions where a pump, valve or elbow is directly upstream or the upstream piping has no straightening element, a longer straight run may be required. Proper filtration upstream of meter A meter installation should always have some type of filtration upstream of the flowmeter. The size of the filter is dependent on meter size and bearing type. Typical recommended filtration requirements for our flowmeters is 10 to 100 micron. Flow Pulsations Pulsations may cause the meter to read inaccurately and excessive pulsations could cause bearing damage. Piping and system components should be arranged to keep pulsations below 10% of current flow rate. Bypass piping Required for installation or removal of meter run without the process being shut down. Control valve downstream of meter run Valving downstream of meter run reduces possibility of hydraulic shock. Purge valve Liquid meter runs should be purged of all entrained gas prior to start up to reduce possibility of hydraulic shock. Orientation Orientation of the meter will affect the load on the rotor bearing. For optimum accuracy the meter run should be installed in the same orientation as it was calibrated. Standard calibration orientation is with the meter axis horizontal.

Annexure – II UVC THEORY Viscosity Effects on Turbine Flowmeters Introduction Like any transducer, a turbine flowmeter is sensitive to physical parameters other than the one which is of interest. While designed to measure flow, a turbine meter responds to the viscosity of a fluid as well as its velocity. Following is a brief discussion of the viscosity sensitivity of a turbine flowmeter. The discussion begins with a review of the nature of viscosity and the means by which it is measured. The characteristics of viscous drag and its action within the turbine meter are also discussed. Following this, there is a description of Universal Viscosity Calibration techniques as a practical means of dealing with the viscosity sensitivity of turbine meters. Readers familiar with the viscous properties of fluids may wish to begin with the section on Universal Viscosity Calibration. Nature of Absolute Viscosity Absolute viscosity is the characteristic of a fluid which causes it to resist flow. The higher the numerical value of absolute viscosity assigned to a fluid, the greater the resistance that fluid offers to flow. Water and gasoline are relatively low viscosity fluids which flow very easily. Such a low viscosity fluid is frequently referred to as an approximately inviscid fluid. Motor oil and maple syrup are examples of more viscous fluids which offer far greater resistance to flow. Castor oil and molasses are even more viscous. It is the viscosity of the molasses which prevents it from running out of the bottle when cold. The viscosity of a fluid causes a loss in pressure as it flows, so that an increase in viscosity requires an increased amount of energy to pump fluid at the same rate. When it is necessary to cause fluids to flow through small openings, such as in low capacity flowmeters, too high a viscosity can cause so much pressure loss that it becomes impossible to establish the desired flow rate. Expressed another way, flow from a constant pressure source will decrease as the viscosity of the flowing fluid increases. The viscosity of a liquid is highly temperature dependent. An increase in temperature will cause a decrease in viscosity. For this reason, it is possible for a temperature change to affect the performance of a flowmeter considerably.

Absolute Viscosity Defined The absolute viscosity of a fluid is a measure of its resistance to shear stresses. Shear stresses occur whenever a fluid experiences an angular deformation as it flows. Consider a fluid flowing over a surface as illustrated in Figure 1. Due to viscosity, the film of fluid next to the surface will be sticking to it and will, therefore, have zero velocity. Fluid further away from the surface will slip over the fluid beneath it as it moves to the right. Since each successive layer of fluid will slip over the layer below it, the velocity of the fluid will increase with the distance from the surface over which the fluid is flowing.

Consider then a small cube of fluid as shown at point A in Figure 1. By the time this small cube of fluid gets to B it will have to have deformed as illustrated because the fluid further from the surface is moving faster. This is referred to as angular deformation. Now consider the forces acting between such a cube and a similar cube directly above or below it. Since the two cubes are in effect sliding over each other, there is a force between them resisting this sliding or shearing. If H is the height of the cubes, and V is the difference in velocity between the center of the top cube and the center of the bottom cube, the force between be represented by

In differential form, this equality is usually expressed as:

assuming that y designates vertical distance. Fluids are said to be Newtonian fluids if the proportionality factor, µ, is a constant. The constant of proportionality is defined as the absolute viscosity of the fluid. It is usually identified by the Greek letter mu. Kinematic and Absolute Viscosity The ratio between shear forces and the velocity gradient in a fluid (as defined in the previous section) is absolute viscosity, and is identified with the symbol µ (Greek letter mu). Because the ratio for absolute viscosity to density appears in so many engineering equations, a second symbol is assigned to that ratio. The symbol used is v (Greek letter nu). Thus:

It is the kinematic viscosity, v, which is of interest in turbine flowmeter applications. Thus, both the absolute viscosity and the density must be known. Knowing the specific gravity is equivalent to knowing the density. In the metric system, µ is expressed in centipoise, v is expressed in centistokes, and v = µ/S, where S is the specific gravity. Viscosity Measurement Viscosity is most commonly measured with rotating cylinder viscometers or capillary tube devices. Rotating cylinder viscometers measure the shear forces associated with a fluid, and thus determine the absolute viscosity. The absolute viscosity is usually determined in metric units, which are centipoise. The density of the fluid must then be measured so that the kinematic viscosity may be obtained. Kinematic viscosity is expressed in centistokes in the metric system. When absolute viscosity is expressed in centipoise and density is expressed in gram/cc, the ratio will result in centistokes. Capillary tube viscometers measure kinematic viscosity directly. This is done by observing the time required for a specific volume of fluid to flow through the capillary tube. Rather than converting the data to centistokes, the raw data itself has been used to establish some viscosity scales, and they then have units in seconds. The Saybolt Seconds Universal (SSU) is an example of such a viscosity scale. Temperature & Pressure The absolute viscosity of a fluid is strongly influenced by temperature. As temperature increases, the viscosity of a liquid decreases and the viscosity of a gas increases. It is customary to express these relationships as a plot of viscosity vs. temperature, and such plots can be found in many references for common engineering fluids. Many oils have a straight line characteristic if the viscosity temperature relationship is plotted as on an ASTM chart (ASTM D-341-43). An example of such a plot is shown in Figure 2 for MIL C-7024B, Type II solvent and MIL H-5606-B Hydraulic Fluid. Because of the temperature dependency of viscosity, it is important to specify the temperature as well as the service medium for turbine meter systems. The influence of pressure on absolute viscosity is usually neglected, and this approximation is reasonable for low pressures. However, for pressures over about 1000 PSI, the absolute viscosity of a fluid may be a strong function of pressure. Generally, an increase in pressure will increase the viscosity of a liquid. Kinematic viscosity is the ratio of absolute viscosity and density. Therefore, if density changes with temperature or pressure, the kinematic viscosity will also change. For gas applications, kinematic viscosity is a strong function of pressure. Remember kinematic viscosity is the key fluid parameter that influences the turbine meter performance.

Laminar & Turbulent Flow Fluid flow is characterized as being either laminar or turbulent. In laminar flow the fluid moves in layers, with one sliding smoothly over the other. There is no mixing of fluid from layer to layer, since viscous shear forces damp out relative motions between layers. Since each layer of fluid is in effect flowing over the one adjacent to it, the fluid velocity increases with the distance from the pipe wall. The resulting velocity profile is approximately parabolic in shape. This is illustrated in Figure 3.

In turbulent flow, there are no discrete layers of flowing liquid. The momentum of the fluid overcomes the viscous shear forces, and there is extensive and continual mixing across the flow stream. This causes the velocity profile across a pipe to be nearly flat. This is illustrated in Figure 4.

A measure of the laminar or turbulent nature of flow is the Reynolds Number (Re). By definition:

The numerator in the Reynolds Number is directly related to the momentum possessed by the fluid. The denominator is the absolute viscosity of the fluid, and is therefore, directly related to the shear forces existing in the fluid. The Reynolds Number is, therefore, a ratio of momentum to viscous forces. Recalling that a predominance of momentum is associated with turbulent flow and a predominance of viscous forces are associated with laminar flow, it is then to be expected that a large Reynolds Number will be associated with turbulent flow. This is because inertial forces associated with the momentum of the fluid will be large in comparison to any viscous shear forces which tend to damp out turbulent motions, and considerable mixing of the flowing fluid will result. Conversely, a low Reynolds Number is associated with laminar flow. The transition from laminar to turbulent flow generally occurs at a Reynolds Number between 2000 and 4000. Another means of visualizing the effects of laminar and turbulent flow is by the nature of the boundary layer. The boundary layer is that layer which is influenced by adherence of the fluid to the pipe wall. Flow essentially moves in layers in the boundary layer, and the flow is laminar. In completely laminar flow, the boundary layers meet in the center of the pipe. However, in very turbulent flow the boundary layers are very thin regions adjacent to the pipe wall. This is illustrated in Figure 5. The more viscous a fluid, the more the shear forces damp out mixing between layers. This allows the fricional drag from the pipe wall to influence boundaries closer to the center of the pipe.

Viscous Drag When a viscous fluid flows over a solid surface, a force is exerted on the surface in a tangential direction. In effect, the moving fluid is attempting to drag the solid surface along with it. The magnitude of this force is dependent upon the viscosity and velocity of the fluid. Experimentally measured drag forces are generally plotted in terms of CD vs. Re on a log-log chart. CD is the drag coefficient and Re is the Reynolds Number. The drag coefficient is defined by:

A plot of CD vs. Re usually looks something like Figure 6. The drag coefficient decreases rapidly with the Reynolds Number in laminar flow, rises abruptly in the transition region, then levels off and eventually decreases slowly in the turbulent region.

Viscous Drag In A Turbine Flowmeter The viscous drag exerted by the metered fluid acts on all of the moving surfaces of a turbine flowmeter. This drag acts within the bearing and in the space between the rotor blade tips and the housing. The viscous drag exerted on the surfaces of the rotor blades produces both a downstream thrust and a retarding torque on the rotor. Because of the viscous retarding forces on the rotor, it does not spin as fast as it would in an inviscid fluid. The rotor actually slips in the stream of flowing fluid, so that the surface of the blades slightly deflect the fluid. The deflection of the fluid by the rotor blades causes a change in momentum in the fluid and a resulting force on the rotor blades. This force on the blades causes a driving torque which overcomes the retarding torque caused by viscous drag forces. The rotor then spins at a rate at which the slip causes the driving torques to exactly cancel the retarding torques. The amount of slip of the rotor will depend upon both the kinematic viscosity and the velocity of the fluid, just as do the viscous drag forces discussed in the previous section. Therefore, the performance of the meter is a function of the Reynolds Number associated with the flow. The performance of the turbine meter does not actually change in exactly the same manner as the viscous drag forces change. This is because the changes in Reynolds Number changes the velocity profile of the flow and this also affects the speed of rotation of the rotor. The viscous drag also contributes to the pressure drop across the turbine meter. Very high viscosities will limit the maximum possible flow rate because of this Universal Viscosity Calibration The calibration of a turbine flowmeter consists of observing the output frequency of the meter for specific rates of flow as determined by the particular calibrator being used. The result is a linear curve as illustrated in Figure 7. However, this curve is valid only for fluids with a kinematic viscosity similar to that of the fluid used in calibration. For operation in high viscosity fluids, the curve in Figure 7 will have less slope and a positive zero offset along the horizontal axis. This is illustrated in Figure 8. Since a different curve will result for every viscosity, this is not a usable form for the calibration data except for single and constant viscosity operation.

A more usable form for the calibration data is called a universal viscosity curve. This is a semi log plot of the sensitivity of the meter as a function of the ratio of the output frequency to the kinematic viscosity. An example is shown in Figure 9. The sensitivity of the meter is commonly known as the K factor, and it is the number of pulses the meter will produce for each gallon of fluid which flows through it. It is derived from the calibration data as follows: K = HZ x 60 / GPM The rational for using the ratio HZ/v may be seen by observing that it is directly proportional to the Reynolds Number for the flow through the meter. Hence the Universal Viscosity Curve is essentially a plot of meter sensitivity vs. Reynolds Number. As such, it reflects the combined effects of velocity, density and absolute viscosity acting on the meter. The latter two are combined into a single parameter by using kinematic

viscosity (v). The Universal Viscosity Curve is formed by plotting K vs. HZ/v for every calibration data point. Typically, thirty points are used; ten each for three different fluids. The thirty points are plotted on a common graph to form a smooth curve. Once this is done, the K factor may be determined for any flow rate in fluid of any viscosity as long as the ratio HZ/v is within the range of values covered by the graph. To determine the flow rate from measured output frequencies and viscosities simply follow the steps shown in Figure 10.

1. Determine output frequency Hz 2. Measure kinematic viscosity v or measure temperature and use temperature to

determine v 3. Calculate Hz/v 4. Read up from known Hz/v to curve 5. Read over from curve to find K factor 6. Calculate GPM: GPM = Hz x 60/K

Annexure III

Operational features of the multiviscous calibration facility at FCRI

Operating Pressure: 10 bar Flow range Uncertainty by

mass Uncertainty by volume

Uncertainty by volume flow rate

300 lpm 0.024% 0.05% 0.075% Turbine flow meter The turbine flow meter is a device, which gives an electrical pulse output proportional to the flow volume passing through the meter. Turbine meters have been used for the flow measurement since 1940’s. When appropriately applied, they offer high accuracy and long service life over a wide range of clean products and operating conditions. Since publication of API Standard 2534 "Measurement of Liquid Hydrocarbon by Turbine Meter Systems" in March, 1970, the turbine meter has gained broad acceptance for custody transfer of petroleum liquids such as liquefied petroleum gases (LPG's), light distillates, and light crude oils, primarily at large petroleum storage and transfer terminals. This flow meter has also been widely used in fluid power applications.

Method of Calibration at the multiviscous oil flow calibration facility “Standing start and stop gravimetric method:” is generally preferred for quantity meters. In this method the flow through the meter is accelerated as quickly as possible from rest to the full test flow rate. At the end of the test, the flow is rapidly stopped; provided the test time is sufficiently long in comparison with the acceleration and deceleration periods the accuracy of the method is not affected. The ON_OFF valve in the system will kept open and the flow through the meter will be adjusted to the required flow rate by controlling the downstream control valve. Once the flow rate is adjusted as per the requirement, the ON_OFF valve will be closed .The initial mass of weigh tank is noted. Then the ON_OFF valve will be opened to allow flow through the meter. The flow will be maintained through the meter for a known interval of time. The time of flow is determined using a limit switch attached to the ON-OFF valve. The mass of oil collected, the time of collection, the temperature of oil, the indicated flow rate .etc. are noted. The density of oil at the line temperature is determined from the on line densitometer. Experiment Aim of the experiment was to evaluate the performance of the turbine flow meter in different viscosity oils viz 1 cst,2cst,100,cst,460 cst oils . The experiment was designed to determine the meter factor of the flow meter at different flow rates. The meter factor is defined as the pulses produced by the meter for unit volume .It was conducted on a 40mm turbine flow meter at ambient conditions. Sufficient upstream and downstream straight lengths are provided for the flow meter to eliminate any disturbance or swirls produced by the piping configurations. Piping configurations are kept same for all the tests .Sufficient proving runs were done to validate the results. After obtaining the data, universal viscosity curve was drawn. The universal viscosity curve(UVC) consists of data from a series of calibrations at various kinematic viscosities covering the range of interest. The data from all the calibrations is presented as k factor as a function of frequency / kinematic viscosity, a continuum of the data will follow a single line. Fig 2. shows a typical calibration curve for a turbine flow meter at 1 cst viscosity. The data was plotted as K factor vs Frequency/viscosity .Super imposing second , third and fourth calibrations results of 2 st,100cst and 460 cst viscosity oils on the above graph produced the universal viscosity curve Note. For most of the flow meter the normal turn down ratio is 10:1.It is a common error to say the turn down ratio as 1: 1000 or above by looking into universal viscosity curve. Extended ranges of calibration at various viscosities that will not follow the universal

viscosity curve concept are called “multiple viscosity calibrations” and are valid only at the discrete viscosities used in the calibration. Conclusion

1. It may be safely concluded that the turbine meter is sensitive to viscosity of the flow media.

2. A drastic variation in the meter factor was observed as the viscosity of the oil changed. From figure 1, it is evident that, the K-factor of the flow meter is inversely proportional to the viscosity of the flow media. This phenomenon is predominant at lower flow rates as the effect of viscosity is more at lower velocities. The calibration results in lighter oils 1cst and 2 cst oil was agreeable within +/-0.5 % However the percentage deviation in meter factor obtained between 1 cst and 100cst oil calibration was within +/- 8 % and for 1 cst and 460 cst is 29 % ,which implies the need of multiviscous calibration .

3. Universal viscosity curve can be represented by a polynomial equation which can be programmed into a flow computer. The computer receives a temperature sensor input (to calculate viscosity based upon a given equation) and the flow sensor input and then calculates out the UVC polynomial equation to give viscosity corrected flow rate.Few suggestions to minimize the effect of viscosity on turbine flow meter performance are

4. Perform the calibration in at the same fluid or with a fluid having same kinematic viscosity at operating temperature. No corrections to the data are required when using this procedure .

5. Liquids that have viscosities below 0.6 cst that don’t have a vapour pressure above atmospheric ,special calibration systems have to be implemented to maintain the fluid in the liquid state.

6. Liquid meters may be calibrated in high pressure air also, if the kinematic viscosity matches.

7. If an alternate fluid of equivalent kinematic viscosity cannot be found, then calibrate the meter in which the kinematic viscosity is close to the desired viscosity. The frequency range of the calibration may be adjusted to compensate for the difference in viscosities.

Reference.

• ISA RP 31.1.Recommended practice for specification, Installation and calibration of turbine flow meters.Flow measurement:

• Practical Guides for measurement and control by Mr. D.W. Spitzer

Table I - Calibration result of 40 mm turbine flowmeter in different viscosity oils

Flow rate m^3/hr

A 1 cst pulses/litre

B 2 cst pulses/litre

% variation in k factor (B-A)/A*100

C 100 cst pulses/litre

% variation in k factor (C-A)/A*100

D 460 cst pulses/litre

% variation in k factor (D-A)/A*100

5 168.95 168.67 -0.16 155.95 -7.69 120.22 -28.84 10 170.25 169.44 -0.48 162.71 -4.43 144.46 -15.15 15 170.56 169.92 -0.38 168.21 -1.38 158.03 -7.35 20 169.87 170.10 0.14 172.44 1.51 160.93 -5.26

CALIBRATION RESULTS OF A TURBINE FLOWMETER AT FCRI IN DIFFERENT VISCOSITIES

110

120

130

140

150

160

170

180

0 5 10 15 20FLOWRATE (M^3/HR)

MET

ER F

AC

TOR

( PU

LSES

/LIT

RE)

1 cst 2 cst 100 cst 460 cst