AGA Issues Revised Reports For Orifice Meter Installations.
These simple meters use a plate with a hole in it to partially block the flow through a pipeline, with a pressure sensor measuring the differential pressure across the plate. There are no moving parts and with nine decades of experience, the mathematics of calculating the instantaneous flow rate as a function of the pressure differentials and thus the volume of flowing gas is pretty well understood.
What is not entirely understood are the various effects of certain installations of orifice meters that make a difference in the flow rate measurement compared to the true flow rate.
Until recently, this difference was considered to be "measurement uncertainty." That means the errors in the measurement could be positive or negative.
After years of ongoing tests and data collection, the American Gas Association (AGA) collaborated with the American Petroleum Institute in preparing a revised version of AGA Report No. 3, Part 2, which is used as a reference standard for measurement of natural gas and other related hydrocarbon fluids by orifice meters.
The revised report (also known as API MPMS Chapter 14 Section 3 Part 2 and GPA 8185-00), was published in April 2000. Some of the effects of installation on measurement that were previously assumed as "measurement uncertainty" were identified as bias error while some are still recognized as measurement uncertainty.
The most critical use of meters is at points where custody of the gas changes hands. Thus, the meter becomes a cash register, and it is this financial aspect of metering that drives efforts to perfect the system. One reason for promoting awareness of the revised AGA Report is to make gas sellers and buyers know that there are possibilities of bias error that could be rectified by conforming to the recommendations in AGA 3.
In the United States alone, the flow of natural gas is about 60 Bcf/d. Traveling from the wellhead to the burner tip, molecules of natural gas could easily flow through five meters at custody-exchange points, which multiplies by five the volume of natural gas subject to custody-exchange measurement. Given that different installations can bias the measurement in different directions, it's possible that a buyer-seller in the middle of the natural gas journey could lose both ways -- receiving less than paid for from the seller, and transferring more gas than a downstream buyer pays for.
If one out of five meter installations under-estimates the flow by two-tenths of one percent, the measurement error for 60 Bcf/d will be of the order of 120 MMcf/d. At $3 per thousand cubic feet, estimated loss is about $130 million per year. Obviously, losses could be dramatically higher with higher commodity costs and/or a higher level of measurement error.
For commodity sellers, such as producers, the failure to test meters and comply with the new report could leave them short-changed. For transporters, errors in measurement may be disguised because they fall within the shrinkage tolerances for pipelines and thus could be chalked up to normal operations in the "lost and unaccounted for" category.
The problem is growing larger as natural gas consumption increases in the United States, as the price of that gas increases, as more players enter the market in an unbundled competitive environment and as the "handshake deals" of the old natural gas economy disappear and are replaced by quicker, less-personal deals coupled with litigation to resolve issues.
The new AGA standard applies to new installations, of course, but also comes into play if an old meter tube is refurbished or relocated.
If a meter is tested and found to have a bias, a seller and buyer using that transfer point can agree contractually to make financial rather than physical adjustments if the cost of fixing the problem is deemed prohibitive. But meter owners must keep copies of meter tests and as new players come into the market at that meter site, they can do audits of the meter tests. This is being done with increasing frequency and, in fact, an AGA Task Group has just been formed to prepare a technical report on Measurement Audit Trail.
To conform to the requirements of the AGA Report, in some cases, lengthening the straight pipe upstream of the meter by four times the lengths recommended by prior standards may be difficult if not impossible. An alternative might be to install a flow conditioner, typically a 19-tube bundle of small-diameter pipes in the pipeline. However, in production-area pipelines where wet gas is prevalent, the bundle of smaller tubes can become clogged. In addition, the use of flow conditioners prohibits the use of pigs to clean or to inspect the pipelines, which is a pipeline safety issue.
Another fix to a meter-installation bias might be to reduce the beta ratio, which is the ratio of the orifice hole to the diameter of the pipeline. The maximum beta ratio allowed by the standard is 0.75, meaning that the orifice hole is three-fourths the diameter of the pipeline. Reducing the beta ratio in the meter will reduce the bias in measurement, but it also increases the pressure across the plate (delta P) for the same flow rate and results in a higher permanent pressure loss. This means that more horsepower at compressor stations will be needed to keep the gas flowing. Most pipeline operators have the horsepower available, but boosting it raises the cost of operation.
Still, lowering the beta ratio has some benefits in increased accuracy. To reduce measurement uncertainty, many natural gas producers limit the use of orifice meters to a maximum beta ratio of 0.6 for all of their custody-transfer points. But for certain installations, restricting the beta ratio to 0.6 may still fail to conform to the requirements of the new report.
The revisions in April 2000 were to AGA Report 3, which has undergone several changes since gas industry experts organized in 1924 to investigate on improving gas-flow measurement using orifice meters. The previous revision was released in four parts in 1990-92, but even then it was known that more research, involving considerable expense, was needed on the specification and installation portion of the report.
The AGA 3, Part 2, revision constitutes the fourth edition of the report on Orifice Metering of Natural Gas and Other Related Hydrocarbon Fluids--Specification and Installation Requirements. The AGA calls the changes "major" and says, "conformance to the revised report will avoid possible bias errors in measurement."
Listed below are some of the major changes in AGA Report 3, Part 2:
* UpStream straight pipe length requirements for different piping installations have increased, sometimes by as much as 2.5 to 4 times the installed length required by the previous edition of AGA 3.
* The minimum straight length requirements for orifice meter installations are tabulated in the report as a function of beta ratio, upstream fitting and combinations of upstream fittings.
* Further, the values are separated into those for installations with and without flow conditioners.
* For installations with combinations of upstream fittings, the distance between them dictates the required length.
* New restrictions on flow conditioners are imposed and a specific design of 19-tube bundle flow conditioners is allowed by the report. Most of the prior designs of a 19-tube bundle, including a hexagonal design, are not accepted by the new report. Any flow conditioner that does not conform to the 19 tube-bundle flow conditioner of the report must be flow tested to demonstrate acceptable performance.
* Use of other flow conditioners is allowed, provided that they can demonstrate the performance based on test procedures defined in the new report. This includes the use of flow conditioners with shorter upstream straight lengths than the length defined in the report, if the test results show measurement accuracy within the acceptable limits.
* The maximum allowable differential pressure across the orifice plate in existing fittings can be increased by installing thicker orifice plates in fittings designed for a different plate thickness, provided that the mechanical tolerances conform to the limits defined in the report. For some line sizes, the recommended orifice plate thickness has changed.
* If the upstream straight pipe length in existing installations does not conform to the recommended length in the revised report, a specific design of a 19-tube bundle flow conditioner can be installed at a certain location upstream of the orifice plate.
* The limits of the maximum and minimum pipe wall roughness of the meter tube have changed and are functions of the beta ratio and the nominal pipe size of the meter.
* Results of an in-situ calibration or hydrodynamically similar test performed in a laboratory can be used for flow rate calculation through an existing installation that does not conform to the specifications of the report.
* The minimum limit of eccentricity for certain line sizes and beta-ratios is now removed. The allowable Eccentricity Equation is the same for all line sizes and beta ratios.
* Installation of thermowell upstream is now allowed provided a flow conditioner is used, thus the meter could be used for bi-directional measurement with proper orientation of orifice plate.
The first step toward eliminating bias caused by improper meter installation is to obtain copies of AGA Report No. 3 Part 2. Copies can be ordered by calling (301) 617-7819 and requesting Catalog No. XQ0002. Online orders can be made at www.aga.org; click on Publications, then Resource Catalog, then Measurement. Discounts are available for purchases of six or more copies.
Ali Quraishi is director of engineering services for the American Gas Association in Washington, D.C.
Dr. Zaki Husain is staff consultant to the general engineering department of Texaco Inc. in Bellaire, Texas.