NEW ORIFICE METER STANDARDS IMPROVE GAS CALCULATIONS

Raymond G. Teyssandier
Texaco Inc. Houston
Ronald Beaty
Amoco Production Co. Houston

Gas volume calculations with the new orifice meter standard will be more accurate than volumes calculated by previous U.S. or international standards. The greatest impact is likely to be on unprocessed gas.

The standard is expected to improve the uncertainty range by 0.1-0.5%.

Because of the numerous changes in the document, implementation of all parts will not occur immediately but in a timely manner.

The new American Petroleum Institute, American Gas Association, and Gas Processors Association orifice metering standard is based on the latest research available to the industry and represents the first major change in U.S. calculation procedures since first published in 1935.

On a worldwide basis, the International Standards Organization Technical Committee 28 has recognized this document as the most up-to-date standard on the orifice metering of natural gas.

The changes were based on the latest available data gathered by various U.S. and European researchers.

The new orifice measurement standard is in four parts that separate out the text to facilitate use and simplify updates.

The sections are:

Part 1-General equations and uncertainty guidelines
Part 2-Specifications and installation requirements
Part 3 -Natural gas fluid measurement
Part 4-Implementation.

Parts 1 and 3 outline the fundamental equations for flow calculations. Part 1 is general and can be used for any fluid.

This article addresses the calculations found in Part 3.

DISCHARGE COEFFICIENT

M.J. Reader-Harris from the U.K. and J.E. Gallagher from the U.S developed, based on more than 10,000 data points, the new equation for CD (see nomenclature box). The equation follows the form, used in ISO-5167, first proposed by Stolz of France.

The equation can calculate the coefficient of discharge for corner taps, CD(CT), flanged taps, CD(FT), and radius taps, CD(RT).

The equation's new form requires an iterative solution for the discharge coefficient. One approach is presented in Part 4 of the standard. But for most natural gas operations, other methods such as simple iteration can achieve closure, within the 50 ppm requirement, in two to five iterations.

Unlike the older Buckingham equations, the new equations for flange taps (Equations 1-10 in equation box) do not require the series of steps to ensure that imaginary numbers are avoided. Users are cautioned, however, that two breakpoints are in the equation. These breakpoints approximate physical phenomena in orifice meters. Programmers must take these two points into account.

The first breakpoint is for pipe diameters of less than 2.8 in. This point accounts for the interaction of the pressure field in smaller pipes where there is a large-diameter pressure tap relative to pipe size

The second breakpoint is a quasi laminar-turbulent transition that occurs at Reynolds numbers below about 3,500.

The new calculation, unlike the old, does not include Fr, a noniterative approximation (albeit an excellent one) for the Reynolds number correction.

For those using the ISO equation developed by Stolz, the iteration scheme should be the same.

EXPANSION FACTOR

Regardless of the flow equation's form, it is necessary to calculate the expansion factor. This term is essentially a thermodynamic process factor relating upstream and downstream density.

In the new document the upstream factor, Y1, is unchanged, but the downstream factor, Y2, now includes real gas effects.

Because virtually all meters in the U.S. use the downstream term, this change is very important to the U.S. natural gas industry.

Equations 11 and 12 calculate Y1.

In the old document the relationship between the upstream and the downstream expansion factor was calculated by Equation 13. The new relationship is Equation 14.

The difference is that Equation 14 requires a compressibility factor for both the upstream and the downstream pressure if the downstream pressure is used.

Because two compressibility values are required, the new downstream calculation may have higher uncertainty and take longer to compute.

There are two options to overcome this difficulty.

The first is simply to move all static pressure measurements to the upstream side. Because the standards developing committee recognized that this would be a major problem, a more simple less costly approach is detailed in the standard.

This second method is to calculate the upstream static pressure by adding the differential pressure to the normally measured downstream static pressure. Note that the units must be consistent.

After calculating the upstream pressure with Equation 15, the expansion factor then can be obtained with the same equation used in the past.

With this method, programmers need to ensure that all parameters are referenced to the upstream location. That is, even though downstream pressure is measured, the calculated upstream pressure must be included in the flow equation and for calculating the compressibility factor.

An inconsistent approach can lead to significant errors.

DIAMETER CHANGES

The new standard accounts for variations of the orifice diameter and the pipe diameter caused by temperature change.

In the older version, only the orifice diameter variation with temperature was calculated.

Use of traditional materials will result in changes to both diameters and also the beta ratio (Equation 16). The beta ratio will remain constant only if the plate and the pipe are of the same material.

Equations 17 and 18 are similar to those found in previous editions except that the equations are for linear rather than for area changes.

Note that, as in the past, the temperature at which the orifice plate and the meter tube are measured must be recorded so that the reference diameter (the diameter normally stamped on the plate or flange or fitting) can be determined.

The reference temperature remains at 68 F. for both the orifice and the meter run diameter.

The recording of the "as measured" values also has not changed and remains at 0.001 in.

FLOW EQUATION

The form of the flow equation that is in Part 3 is different from the equation commonly used in the past document.

The new form has, in fact, returned to its original form that is much clearer. Several variants are found in the text of the document. Two are shown as Equations 19 and 20.

Equation 21 is the hourly volumetric flow rate at standard conditions.

The new standard shows all volumes calculated to standard base conditions. Equation 22 converts the volume to other temperature or pressure bases.

To some, these equations may look more complicated than the previously used factor equation. However, the older equation required numerous background calculations to actually calculate the specific factors for the flowing gas.

These factor equations were developed in the age of slide rules or mechanical calculators. With this older technology, taking a square root or raising numbers to odd powers was very involved. Therefore, the factor equation was a very good equation. But the equation in the factor form could sometimes be misinterpreted.

As an example, Equation 23 identifies the term C' as the "orifice flow constant."

For technology using charts, this assumption could be considered true. In most cases, chart processing renders only one number for the differential pressure times the flowing pressure, one for static pressure, and one for temperature. Thus, for a given chart, the above factors are calculated only once.

With electronic flow metering, "real time" calculations are possible. Calculations are performed routinely as fast as once every second. Thus, all of the relevant flow parameters are calculated, essentially, as fast as they vary.

Because the new equation does not lend itself to any form of tabular documentation, field calculations will require a hand held, lap top, or notebook-type computer.

In the office, calculations can be made on anything from a mainframe to a nonprogrammable scientific calculator, if one has enough time. With today's availability of computers, there is no comparison to the trauma felt in 1935 when AGA Report No. 2 required parameters to the power of 5/2 in the Buckingham equation.

COMPRESSIBILITY FACTOR

For natural gases the document reaffirms the use of the latest revision of AGA Report No. 8.

It is important to note that if Part 3 had retained the old flow equation but continued its support of AGA Report No. 8, field calculations would still require the same hardware as for the new equations.

EXAMPLE

In the example box a simple iteration technique is used to solve for CD(FT).

The equations have two unknown components: CD(FT) and the Reynolds number associated with the pipe diameters and flow rate (ReD).

The Reynolds number can be calculated with Equation 24.

Because CD(FT) is in the equation for Qv and Qv is used in the ReD equation, the method assumes an initial value for CD(FT). A CD(FT) equal to 0.6 is an excellent first guess for a natural gas application.

The assumed CD(FT) is used to calculate a first pass Qv. This Qv is then used to calculate an initial ReD. In turn, this ReD is used to obtain a second CD(FT).

This cycle is continued until a CD(FT) is obtained that agrees with the previous CD(FT) through six significant figures.

The iterative solution can also begin by assuming a value for ReD. The same solution for CD(FT) will be obtained. However, up to ten passes may be required for the desired agreement.

Editor's Note: The new standards (ANSI/API 2530-1992) may be obtained from either the American Petroleum Institute or the American Gas Association at a nominal cost of $50/section.