Coal, Gas

Metering Matters – Inaccurate Gas Measurement Can Cost Thousands

Issue 4 and Volume 106.

By: Edgar B. Bowles, Jr.,
Southwest Research Institute

Gas-fired power plants consume large volumes of gas. Accurate measurement of natural gas consumption in gas-fired power plants is a necessary element in commodity pricing of electricity. An accurate measure of the gas consumption is also critical to determining plant efficiency. Inaccurate gas measurement can result in inequitable charges or can give a false indication of a problem with a plant’s operational efficiency.

The method of measuring natural gas transfers is normally specified in the contract between the buyer (i.e., the power plant operator) and the seller (i.e., the gas supplier). An inaccurate custody transfer meter can be a detriment to either the buyer or the supplier of the gas, depending on whether the meter over-registers or under-registers the volume of gas involved in the transaction. One way a power plant operator can identify when an error in the gas measurement may exist is by placing a check meter in series with the gas supplier’s sales meter. Both the sales meter and the check meter must provide precise measurements for the check meter to be of any value in helping identify measurement errors.

This article describes gas metering technologies and the importance of meter calibration. Proper meter selection, installation, operation and maintenance are vital to consistent, accurate gas measurement. Industry standards for various gas metering technologies are available and can be referenced in gas sales contracts to help ensure accurate, equitable gas measurement. The difference between an accurate and an inaccurate flow meter reading can mean significant cost savings to plant operators as well as the gas supplier. Often, offsite meter calibrations are a preferred option for establishing or confirming meter accuracy because test conditions can be more precisely controlled than with an in-situ meter calibration.

A Hot Issue?

Recent deregulation in the gas industry has driven change and innovation in gas measurement technology. Shorter-term contracts and growing spot markets, minimization of operation and maintenance costs, limitation of capital expenditures, and expansion of e-commerce have all influenced gas measurement technology development in recent years. The “hot” issues facing gas measurement are the need for cost reduction, improved accuracy, and more timely availability of measurement data.


Figure 1. Orifice Flow Meter
Click here to enlarge image

Cost Reduction. As the U.S. gas industry continues to consolidate and downsize, gas company resources continue to shrink. Market pressures demand that capital and operating expenses be kept to a minimum, making it more of a struggle to properly maintain gas measurement facilities. This creates a challenge for flow meter manufacturers and meter station designers to provide more cost effective measurement solutions.


Figure 2. Flow Conditioners
Click here to enlarge image

For example, orifice flow meters (Figure 1), which are the most commonly used flow meters for high-volume natural gas measurement, are know to be very sensitive to the upstream piping configuration. Up to 145 diameters of straight pipe may be needed upstream of an orifice meter to ensure that there is no measurement bias introduced by the upstream piping configuration. However, recent research funded by the Gas Technology Institute has demonstrated that the upstream length requirement may be reduced to as little as 10 pipe diameters with the use of a high-performance flow conditioner. Flow conditioners (Figure 2) are devices placed upstream of the meter that eliminate flow field distortions by redistributing the flow across the pipe cross section. This order of magnitude reduction in orifice meter installation length can result in significant cost savings, both in terms of reduced capital cost for the installation and reduced operating cost due to reduced measurement error.

Accuracy. Measurement errors can adversely impact a power company’s bottom line, and also result in customer disputes, litigation and regulatory agency concerns. Flow meter manufacturers and gas industry research organizations are continuing to improve traditional measurement methods and to develop new, more accurate methods. The benefits of improved gas measurement accuracy include proper accounting at custody transfer points, more accurate determination of plant efficiency, and improved customer satisfaction. Just five years ago, volumetric flow rate error levels on the order of ±1 percent were the industry norm. Now, measurement error levels on the order of ±0.25 percent are realistically achievable.

Timely Measurement Data. A decade ago, paper charts or mechanical counters recorded most gas flow rate data. Data acquired via these recording methods were prone to error and typically took weeks or months to process. With the growing popularity of short-term gas contracts and the evolution of e-commerce for the sale of gas, there is a fast-growing demand for real-time measurement data. Measurement equipment manufacturers have responded by making better use of embedded microprocessor technology to help provide gas measurement data in real time, in electronic format. Specific benefits include improved measurement accuracy, timelier transfer/distribution of the measured data, and enhanced self-diagnostic capabilities of the measurement equipment.

Gas Meter Standards

In the U.S., natural gas is bought and sold as a commodity and may change ownership several times from wellhead to burner tip. Standardization of the measurement process at custody transfer points is critical for fair and equitable transactions. Industry standards writing groups, such as the American Petroleum Institute (API), the American Gas Association (AGA), and the Gas Processors Association (GPA) are continually working to update existing gas measurement standards and to create new ones as new measurement technologies make their way into the marketplace. These standards give guidance on the proper use of the various measurement technologies available today and result in industry-wide uniformity in the application of these technologies. For example, AGA Report Numbers 3, 7, and 9 provide guidelines for orifice, turbine and ultrasonic flow meters, respectively.

Click here to enlarge image

Table 1 shows the performance characteristics, industry standard, and power plant application for the most common flow meter types.

Periodic Flow Meter Calibration

It is imperative that the cash register of any business be accurate. It is a cliché, but nevertheless, a fact, that gas custody transfer flow meters are the “cash registers” at the gas delivery points. One very common misconception is that gas measurement errors are random and that these errors will “average out” to zero over the long term. To the contrary, most measurement errors are bias or systematic errors. Bias errors are repeatable in both magnitude and direction. That is, a biased gas meter that under-registers today will under-register tomorrow and the days that follow.

There is no typical range of bias errors that may be expected for gas flow meter installations. Errors of more than 20 percent have been reported. For example, if an orifice meter is used to measure the gas flow, the orifice plate typically has a beveled edge on one side of the bore and a sharp, square edge on the other side. The plate is designed for the square-edged side to be upstream in order for the flow rate measurement to be correct. If the plate is installed backwards, with the beveled edge upstream, approximately a 20 percent bias error (i.e., under-registration of the flow rate) results. There have even been instances reported in which orifice plates have been left out of an orifice meter for extended periods, resulting in no measurement of the gas flow for months at a time! For a reasonably well-designed and maintained meter installation, the bias error will typically be something less than 0.5 percent (and it could produce either an under-registration or over-registration, depending on the circumstances). Fortunately, nearly all measurement biases can be eliminated through proper meter calibration, installation, operation and maintenance.

To illustrate the economic implication of an accurate flow meter calibration, consider the hypothetical case of an 8-inch diameter ultrasonic flow meter serving as a custody transfer meter at a gas-fired power plant. Transmission-grade natural gas is delivered to a power plant at a line pressure of 50 psig. The average gas velocity through the custody transfer meter is 50 feet per second (about the mid-range of a typical meter). Assume that the meter was not flow calibrated prior to installation and was delivered from the manufacturer with a +0.7 percent meter bias (i.e., the maximum meter error allowed in American Gas Association Report No. 9). If the value of the gas being transported is approximately $3.00/MSCF, the +0.7 percent meter bias will result in a $47,000 error in favor of the seller over one year’s time. The current cost of flow calibrating a meter of this size and capacity at a reputable third-party flow lab would be about $7,000. Thus, the cost of a flow calibration to correct for the meter bias could be recovered in less than two months.

Meter Calibration

The larger the volume of gas delivered to a power plant, the more financial risk there is from mis-measurement of the gas. One way of ensuring accurate measurement is to periodically re-calibrate the flow meter. Flow meter re-calibrations can be performed either in-situ, using a reference test flow meter plumbed in series at the meter station, or offsite, at a third-party test facility. There are no gas industry standards for flow meter re-calibrations. Most power plant operators have developed their own guidelines through years of operational experience. In some instances, government regulatory requirements may apply. The time interval between meter re-calibrations is usually determined by the meter station design and operating conditions. For example, devices with moving parts that can wear, such as turbine meters, tend to have the shortest re-calibration interval. Most power plant operators using turbine meters re-calibrate their meters on a one to three year cycle.

For in-situ calibrations, the field meter station must be designed to accommodate the installation of a reference test flow meter to compare to the field meter. Because there are no gas industry standards or guidelines for in-situ or field-meter proving, the test installation configuration and the test protocol are left to the discretion of the parties involved. Offsite meter calibrations are often preferred because test conditions can be more precisely controlled than in the field. This situation typically results in a much more precise (and usually more cost-effective) meter calibration.

If an offsite meter calibration is the preferred option to confirm the performance of a power-plant meter, there are several important factors to consider when selecting a qualified calibration laboratory: responsiveness to client needs; independent, unbiased business perspective; ability to maintain confidentiality; ability to approximate field operating conditions in the test lab; traceability of the laboratory’s reference measurement standard(s) to national and/or international standards(s); documented history of good laboratory measurement accuracy, repeatability, and reproducibility; cost and range of services; experienced staff having the ability to quickly troubleshoot and correct problems; flexibility in scheduling tests; effective and timely reporting and record keeping; good working relationship with gas measurement equipment manufacturers; familiarity with all industry standards for natural gas measurement; and active involvement in developing measurement technology/standards.

A third-party meter calibration laboratory will typically perform a meter calibration over the normal operating range of a meter. The accuracy of the field meter will be evaluated over a range of flow rates (and, sometimes, over a range of line pressures and temperatures) that is representative of field service. The field meter will be compared to the reference flow rate standard of the calibration lab. Reference standards vary from lab to lab, but may include gravimetric weigh tanks, bell provers, critical flow Venturis (also known as sonic nozzles) and turbine meters.

Click here to enlarge image

Meter calibration results are normally provided to the client in tabular and graphical form in a test report. Figure 4 shows an example ultrasonic flow meter calibration plot from the Gas Technology Institute Meter Research Facility located in San Antonio, Texas. In this case, the field meter was tested over a range of flow rates at one line pressure. Multiple measurements were made at each test flow rate to establish the repeatability of the meter. As Figure 3 shows, a correction to the meter calibration factor (for about a 0.75 percent under-registration of the flow rate) was made and then confirmed via a flow rate check. For this meter calibration, the flow rate reference was a bank of critical flow Venturis having a total measurement uncertainty of approximately ±0.15 percent.

Author

Edgar B. Bowles, Jr. is Manager of the Fluid Systems Engineering Section at Southwest Research Institute in San Antonio, Texas. He has worked at Southwest Research Institute for 23 years, involved with a variety of projects for the oil and gas industry. Since 1995, he has managed the Gas Technology Institute Metering Research Facility located at Southwest Research Institute. Bowles holds Bachelor of Science and Master of Science degrees in mechanical engineering from Southern Methodist University.


Nevada Cogen – Check Your Meter

The Garnet Valley Cogeneration Plant, owned and operated by Nevada Cogeneration Associates (NCA), is located approximately 16 miles north of Las Vegas, Nevada. The plant produces electricity for the local utility, Nevada Power Company, and thermal energy for the adjacent Georgia-Pacific wallboard plant. At the site, three GE LM2500 gas turbines and a GE steam turbine produce a nominal 85 MW of power. NCA #1 nominally uses 18,000 MMBtu per day of natural gas for its cogeneration operations. The fuel is transported by the Kern River Gas Transmission Company to a pipeline operated by the Southwest Gas Corporation (SW Gas). The gas is purchased and managed by Texaco Natural Gas. The gas supply is regulated to 425 psig at the plant fence line. Fuel flow to each gas turbine is measured independently in the plant. A SW Gas custody transfer turbine meter at the NCA #1 plant boundary measures the total flow of gas consumed by the plant.

NCA identified a variance between the cumulative fuel flow measured by in-plant instrumentation and the amount reported by the custody transfer meter, which it tracked over a one-year period. The potential error indicated that gas flow to NCA #1 may have over-registered and resulted in an overpayment by NCA in excess of $52,000 per month. As the error appeared to NCA to have exceeded the 2 percent tariff upper limit, SW Gas could have been required to correct the meter and reimburse NCA #1 for over-billing. Alternatively, a cumulative error in the in-plant check meters could have been the source of the discrepancy.

SW Gas made several attempts to assess the actual metering error and determine the cause of the error. After upgrading the meter station and replacing the meter corrector, however, the discrepancy between the custody transfer meter and in-plant meters remained the same. NCA checked the in-plant meters, which were found to be operating within expected accuracy limits. NCA and SW Gas agreed to remove the turbine meter and the corrector for testing by a third party. Unfortunately, due to corrections required from that testing, the inaccuracy of the meter could not be agreed upon.

With the assistance of Texaco Central Engineering, NCA #1 determined that a third party test could be performed at the GTI Metering Research Facility at the Southwest Research Institute (SwRI) in San Antonio, TX. The SwRI facility met NCA’s test accuracy requirements of a nominal ±0.2 percent using pipeline quality natural gas at the NCA supply pressure and temperature.

Following calibration of the meter and evaluation of the corrector at the GTI Metering Research Facility, SwRI found that the combination had over-registered the volume of gas by an average of about 1.088 percent of reading. SwRI calibrated the meter and corrector and returned them to service. Both the custody transfer and check meters were in agreement. Based on an average gas consumption rate of 534 MMSCF/month, NCA estimates that it recovered the cost of the high-accuracy calibration in 20 days.

Since the accuracy level of 1.088 percent was within the 2 percent allowed in the SW Gas tariff, no billing adjustments were required. The calibration enabled NCA to reduce future gas metering charges by more than 1 percent, resulting in an annualized reduction of gas cost (savings) of approximately $205,000.