Вы находитесь на странице: 1из 15

The Americas Workshop 26-28 April 2011

The Effect of Geometry on Differential Pressure Meter Performance


Klaus J Zanker, Letton-Hall Group USA Philip A Lawrence, Cameron Valve & Measurement USA

Introduction

Recently there have been various independent studies regarding the effect of unprocessed hydrocarbon fluids on the geometric stability and application of various types of differential pressure (dP) meters used in upstream flow measurement, particularly in the field of erosion, a key parameter in the operation of meters that obstruct the flow (dP devices). Most of the technical work for this type of erosion testing has been performed in a laboratory using techniques to accelerate years of wear into a short period spanning a few months. This paper discusses and focuses on the relative differences in operation and application between the main dP meter types that have been subject to erosion testing based on this accelerated wear methodology. Techniques together with information showing methods to reduce the wear, particularly in the newer cone meter family, are discussed. 2 The Differential Pressure (dP) Meters

Daniel Bernoulli 1700 1783 and Giovanni Battista Venturi 1746-1822 established the basic theory:

Figure 1. Venturi Meter Theory Conservation of energy in a perfect fluid (incompressible and frictionless) gives:

By continuity the flow Q is the same at sections 1 & 2, hence

The Americas Workshop 26-28 April 2011

These two concepts yield the volume flow:

Where: dP = P1 P2 and = the fluid density Practical Equation To account for viscosity and compressibility two empirical factors are introduced: CD = the discharge coefficient Y = the expansion factor 2 a = the throat area ( D2 /4) 4 -0.5 E = (1 ) = velocity of approach factor

Volume flow

and Mass flow

Note that the fluid density is required to calculate either mass or volume flow. The first practical application of a Venturi meter was made by Clements Herschel in 1887. Up to that time positive displacement meters were accepted but inferential meters were a new concept viewed with suspicion. It was soon realized that any restriction would produce a dP that could be related to the flow rate, which led to a large and ever growing class of dP meters. The recent interest in multiphase/wet gas flow has led to more innovation with dP meters. CD accounts for viscosity and it depends on the Reynolds No Re = V D / . Y accounts for compressibility and it depends on the Mach No M = V / c. For liquids Y = 1. For gases the speed of sound c = (k P/), with k the isentropic exponent.

2.1

The Venturi Meter

Figure 2 - Classical Venturi Meter Cross Section

The Americas Workshop 26-28 April 2011 The Venturi has CD = 0.99, only 1% deviation from the Bernoulli theory. This is because the acceleration from the upstream pipe to the throat tends to produces a uniform velocity profile, with a thin boundary layer blocking just 1% of the throat area. The high throat velocity produces the dP, but it also makes the throat taping sensitive to imperfections and liable to erosion. The standardized un-calibrated Venturi is available with from 0.4 to 0.75 to meet the 1.0% uncertainty of CD. The downstream cone acts as a diffuser to recover some of the kinetic energy at the throat. This makes the Venturi a relatively low loss device, but at the expense of size, weight and manufacturing cost. 2.2 The Orifice Plate

The orifice is a much smaller and lighter device, consisting of a thin plate (1/4 in) held between flanges. The pressure taps are in the flanges and do not experience a high velocity. The main feature of the orifice plate is the sharp edged concentric hole, which makes the flow separate and contract downstream to the Vena Contactor. This is like the Venturi throat, but formed by fluid streamlines and not solid surfaces. The orifice CD = 0.6, which reflects the Vena Contactor area being less that the orifice bore area. The orifice does not have a diffuser and suffers a sudden expansion loss, but despite this it almost displaced the Venturi because of its simplicity in single phase flow. The sharp edge is sensitive to damage and this led to the introduction of a fitting (Daniel 1930) that allows the orifice to be taken out of the line under pressure to examine the sharp edge. The sharp edge is not good with erosion making the orifice less desirable for multiphase flow, but it is used with wet gas. The standardized un-calibrated orifice plates are available with from 0.2 to 0.6 to meet the 0.5% uncertainty of CD. 2.3 The Cone Meter

The cone meter is an annular Venturi without a diffuser and the low pressure tap is in the base of the cone, where it does not experience a high velocity or erosion. With multiphase flow in a horizontal pipe the meter is self cleaning. Liquid and solids will not be trapped at the bottom of the pipe and gas will not be trapped at the top of the pipe. The throat is formed by the annular area between the pipe wall and cone maximum diameter. This physical throat is not the highest velocity location because the angle of the cone directs the flow to contract further downstream, giving CD = 0.82 to 0.85 Figure 3 - Horizontal Cone Meter

The Americas Workshop 26-28 April 2011

With multiphase flow it is common to install the cone meter on a blind Tee, turning the horizontal flow to vertically upward flow to improve mixing and reduce gravity effects. Liquids and solids will tend to fill the low pressure cone base tap, giving a false differential pressure. The solution is to use a wall tap a little downstream of the cone. This wall tap experiences the highest velocity making it sensitive to imperfections and liable to erosion, loosing the advantage of the cone base tap. Slug flow could bend the support strut. The cantilever strut can be replaced by a diametric strut and/or three welded web supports between the downstream cone and pipe wall can give additional structural strength to the meter.

Figure 4 - Vertical Cone Meter

The cone meter is available with from 0.45 to 0.85, but it has not been standardized yet, apart from the API chapter 22.2 D.P test protocol, which details a method for performance type approval. 2.4 The Wedge Meter

Figure 5- Wedge Meter Cross Section The wedge meter was designed as part of the Dual-Stream wet gas meter, which consists of a Venturi in series with the wedge. It was chosen to have a very different response to wet gas (over reading) than the Venturi, so that both the gas and liquid flow could be measured. The wedge is very different to the Venturi, orifice and cone meters, which are axi-symmetric and 2 by varying the area ratio ( ) produce a series of meters. The wedge meter is three dimensional 2 and with the wedges reaching the centreline it only has one area ratio = 0.5 or = 0.707. The wedge meter has a CD = 0.79 and has not been standardized.

The Americas Workshop 26-28 April 2011

The Venturi, orifice and cone meters have a throat with a low pressure tapping at the maximum velocity. The wedge does not have an obvious throat and the low pressure tapping is not at a throat, but in the pipe wall. The wedge meter is closer to a total loss device than a classical Bernoulli device. 3. Meter Comparisons

The Venturi, orifice, cone and wedge meters are different and it is not obvious how to compare them. The discharge coefficients are different as seen in Fig 6.
1.05 1 0.95 0.9 0.85 0.8
Cd

Cone 0.75 0.7 0.65 0.6 0.55 0.2 0.3 0.4 Venturi

Orifice Wedge

0.5 Beta

0.6

0.7

0.8

0.9

Figure 6 - Meter Discharge Coefficient (per beta) Another way of comparing the meters is by the differential pressure they produce for the same flow in the same pipe size (same V1).

and

With J = CD Y E a / A = CD Y E
2

Then we get

1/J is a multiplier that expresses dP in terms of the upstream dynamic pressure (velocity head). Taking an example from Fig 7: a = 0.55 Venturi, a = 0.6 cone and a = 0.68 orifice all 2 produce the same dP (10 V1 / 2).

The Americas Workshop 26-28 April 2011

1000

Cone

Orifice

Venturi

wedge

100

dp / V^2/2

10

1 0.2 0.3 0.4 0.5


Beta

0.6

0.7

0.8

0.9

Figure 7- Meter Differential Pressure If the overall energy loss is important the loss/differential is relevant because it is the loss incurred for the desired differential pressure.
1 0.9 0.8 0.7 0.6
Loss / Differential

0.5 0.4 Cone 0.3 0.2 0.1 0 0.2 0.3 0.4 0.5 Beta 0.6 0.7 0.8 0.9 orifice Venturi Wedge

Figure 8 - Meter Pressure Loss The low loss of the Venturi is due to the diffuser downstream of the throat, but it increases the size, weight and cost of the meter.

The Americas Workshop 26-28 April 2011 4. Geometry

There is a major difference in the geometry of a circular (Venturi and orifice) and annular (cone) 2 throat. For the Venturi and orifice = d/D (with D = pipe diameter, d = throat diameter) and is 2 2 2 2 2 the area ratio. With the cone meter = (D dc ) / D = 1 (dc/D) (where dc = cone diameter). The effective for the cone meter is the square root of the area ratio. The annular gap (D dc)/2 in the cone meter will always be less than the bore d of a circular throat for the same , as seen in Fig 9 below.
1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Annular Gap Circle d Cone dc

Beta

Figure 9 - Meter Geometry for D = 1 We can take the simple numerical example of = 0.6 with the Venturi bore d = 0.6. For the cone 2 2 2 2 meter = 0.6 = 0.36 = 1 - dc giving dc = 1- 0.36 = 0.64 and the cone diameter dc = 0.8. The annular gap = (1 - 0.8)/2 = 0.1. Not only is gap 1/6 the Venturi bore, but the surface area or perimeter is very different. Taking the same example: the Venturi throat perimeter = 0.6 while the cone throat perimeter = (1 + 0.8) or 3 times the Venturi. This is not good for the cone meter in erosive service. It is the surface that is eroded and because of the narrow gap the change in area is more than a Venturi with the same . However, there are some advantages of the cone meter geometry: To measure fluid properties with a gamma density meter a narrow annular gap allows a weaker radio active source to be used and provides a better sample. The cone meter requires short straight pipe entry lengths due to the large surface area and narrow annular gap improving flow conditioning and the increase in radius reducing swirl (conservation of angular momentum). The cone meter can achieve high . With = 0.8 the cone diameter is 0.6 and the annular gap is 0.2. An orifice with 0.8 bore has a 0.1 annular dam, which is very sensitive to flow profile effects and not recommended for accurate measurement (Fig 10).

The Americas Workshop 26-28 April 2011

CONE

0.6

ORIFICE

0.8

0.2

0.1

Figure 10. = 0.8 Cone and Orifice 5. Meter Erosion

There has been experimental and Computational Fluid Dynamic (CFD) work on flow meter erosion. The CFD gives a good insight into the process [Ref 1 & 2].

Figure 11 - Venturi - Predicted Velocity Contours (m/s)

The Americas Workshop 26-28 April 2011

Figure 12 - Venturi - Predicted Erosion Depth Contours (mm)

Figure 13 - Cone - Velocity Contours (m/s)

The Americas Workshop 26-28 April 2011

Figure 14 - Cone - Predicted Erosion Depth Contours (mm)

Figure 15 - Wedge-Predicted Erosion in Water-Sand (kg/m2-s)

There is also experimental data using high sand concentration and high velocity to accelerate the erosion process to yield lab results in a few months instead of many years in the field.

10

The Americas Workshop 26 28 April 2011 26-28 Table 1. Summary of the Erosion Test Conditions. Condition Run Time (hr) Average Sand Concentration (% by wt) Average Temperature (F) Average Venturi DP (psi) Sand/Water Mixture Density (g/cc) Average Venturi Flow (gpm) Mass of Sand Through Meters (lbm) Value 196 2.28 80.4 13.85 1.012 121.5 274,810

Figure 16-Photographs of the Wedge Meter Showing the Eroded Area. 16-Photographs

11

The Americas Workshop 26-28 April 2011

Figure 17- Photographs of the Cone Meter

Figure 18 - Photograph Showing Erosion of the Venturi Throat Taps. Table 2. Summary of the Erosion Test Results.

Meter Venturi Wedge Cone

Beta 0.5 0.7 0.5

% Cd -0.63 1.72 10.77

Max Scar in Location 0.014 bore tap 0.343 wall 0.14 wall/tap

The Venturi is least affected, because the tap and bore erosion counteract each other. The Wedge with the highest and lowest dP has the worst erosion scar (0.343), but the change in CD (1.72%) is modest. This is because the scar is in the pipe wall and not on the wedges. The cone scar is modest (0.14) but it affects the narrow annular gap of the cone throat, causing a large (10.77%) change in CD. The effect of erosion on meter performance is complex. It depends upon , the meter design, the location of the scar, the size of the scar and the sensitivity of the meter to the scar.

12

The Americas Workshop 26-28 April 2011

Probably the best way to compare erosion in different meters is on the basis of equal dP, because dP is the driving force that produces the erosion. 5.1. Meter Erosion Parameters

It has been found that erosion depends on the following factors [Ref 3]:

E c cc .
Where: Ec cc

msand n .k.Vthroat metal

is the maximum increase in the scar depth over the time period (mm) is a constant to account for the meter design and the units

msand is the total mass of sand that has passed through the meter in the time period (kg) Vthroat is the mean fluid velocity in the cone throat (m/s) metal is the density of the material from which the meter throat is constructed (kg/m ) 3 (metal = 7850 kg/m for steel grade materials) k n is a material constant (k = 2 x 10 for steel grade materials) is a material constant (n = 2.6 for steel grade materials)
-9 3

The main factors to reducing erosion are to reduce the amount of sand and Vthroat. Reducing the mass of sand has a direct linear effect on the erosion scar. Reducing Vthroat has the largest effect (V ), so halving the velocity reduces the scar depth (by 2 = 6.06) six fold. Vthroat can be reduces by increasing the meter size or by increasing the meter . This information can be used to reduce an accelerated erosion test to more normal conditions. 5.2 Erosion Resistance (Cone Meter Types)
2.6 2.6

One method to negate the effect of erosion is to fit a sleeve in the rear of the meter body made from a hard-facing material or cobalt based alloy. The hard facing can also be applied to the main parts of the cone differential producer element. This will help to protect and allow the device to be used without worrying about a change in geometry when sand content in the fluid is high because of the hardness of the sleeve (Fig 19a). The hard faced sleeve shown as #132 in the sectional drawing (Fig 19b) is interchangeable from the rear end of the meter body. The cone # 18 and the support cantilever #19 can also be hardfaced to enable a longer life under extreme sand erosion conditions. Hard-facing materials can be applied to any other dP devices when it is known where the erosion occurs. 6. Area Ratio Change.

The use of the anti-erosion sleeve idea can be used to change the beta ratio of an existing meter without changing the diameter of the cone. It is an option that can be used in extreme

13

The Americas Workshop 26-28 April 2011 circumstances where the meter range ability has been affected by a drop in the field flow rates (Fig 20). Some form of calibration to determine the new CD would be needed [Ref 4]. Field calibration of the sleeve system with a test separator is also possible when using the meters on gas dominant full-well stream measurement applications.

Fig 20- Area or Beta-Ratio Changer Using a Sleeve

7.

Conclusions

The effect of erosion on meter performance is complex. It depends upon the application, the , the meter design, the location of the scar, the size of the scar and the sensitivity of the meter to the scar. It is not obvious how to compare the erosion performance for different meter designs, but the 2 same dP for the same flow (1/J ) seems reasonable. Erosion can be reduced by using lower flow velocity, higher beta and avoiding pressure taps in high velocity regions. The relation between erosion, sand mass and velocity allows accelerated erosion testing to be related to practical cases. CFD is a useful tool for predicting where erosion occurs and to estimate its effect on meter performance. Knowledge of the scar location allows the application of hard faced erosion resistant materials.

14

The Americas Workshop 26-28 April 2011 The newer geometry devices that are now available enhance the dP device family, are useful in the more unusual areas of measurement (Upstream, wet gas, sub sea, etc.) and make new developments possible. Recent work with the cone meter using a hard-faced sleeve system to combat sand erosion and the bonus of being able to change the beta ratio, without changing out the meter, is a useful addition.

7.

References

1) BARTON, N.A., ZANKER, K.J. & STOBIE, G. Erosion Effects On Venturi and Cone Meters, ThAW 2010 2) BARTON, N.A., et al. Erosion in Subsea Multiphase Flow Meters, ThAW 2011 3) HAUGEN, K., KVERNVOLD, O., RONOLD, A. & SANDBERG, R. Sand erosion of wearresistant materials: erosion in choke valves. Wear 186-187, pp 179-188, 1995. 4) LAWRENCE/BRAID Cameron Inc (USA & Canada), Cone Equations for Flow Computers a Technical Document - 2006

15

Вам также может понравиться