Proving clamp-on gas ultrasonic flowmeter technology in the field
When a Louisiana-based natural gas producer looked at installing additional points of flow measurement on natural gas lines coming into their facility, they found available technologies were difficult to adapt.
This online extra is a companion to ''Ultrasonic Flowmeters: A 'Sound' Technology'' from Control Engineering , October 2002. Click here to return to the original article .
Panametrics' GC868 flowmeter measures on-shore flow in a 6-in. CS pipe with 900-psig ambient temperature natural gas coming in from ''Platform 1'' in the Gulf of Mexico. The device's Div. 2-certified electronics are shown in the foreground near the bottom of the photograph. The transducer installation is at the top, slightly off-center to the right.
When a Louisiana-based natural gas producer looked at installing additional points of flow measurement on natural gas lines coming into their facility, they found available technologies were difficult to adapt. In the past, orifice meters would have been considered for the application. However, the cost of orifice meters coupled with the fact that they could only be installed during shutdown prevented this from being implemented within a realistic timeframe. Although clamp-on ultrasonic flowmeters would solve the ''no shutdown'' problem, the technology was not considered adaptable to gas flow except at very high pressures, several thousand psig or greater. During their investigation of available technologies, engineering became aware of the relatively newly available GE Panametrics' GC868 clamp-on gas flowmeter, a technology that extended clamp-on ultrasonic technology to gas applications for 3-in. dia. or greater pipes with pressures over 90 psig.
To investigate this new technology, engineers decided to outfit one of several pipelines coming into the plant. The GC868 was able to be easily adapted because the whole installation process could be accomplished without shutting down the process flow or disrupting normal operations, the key factor in the customer's decision to proceed with the project.
Murphy's Law seems to apply!
The first line to be outfitted was a 6-in. CS line with 900-psig natural gas at ambient temperature. After the commissioning, the meter read 11.7-11.8 MMSCFD. Engineers at the site were perplexed by this value because they expected a measurement of approximately 25.3 MMSCFD based on an orifice flowmeter reading at the platform. The platform was located about 20-mi. away in the Gulf of Mexico. Diagnostics on the GC868 flowmeter indicated that it was operating correctly and that it was indicating an accurate volumetric flow measurement. Even so, engineers onsite believed that the GC868 was reading only about half the flow it should have read.
GE Panametrics' engineers returned about a week after the initial commissioning and found that the readings had not changed significantly. Because there was plenty of open space on the pipe at the installation point, the transducers were relocated to rule out quality of transducer installation or pipe anomalies at the original installation site. Subsequent to the relocation, the GC868 flowmeter still indicated 11.7-11.8 MMSCFD and diagnostics indicated the meter was functioning correctly. Detailed discussions followed with engineers on site. A more detailed review of the piping system was done.
It was confirmed that there was a "well head take-off" line upstream of the clamp-on gas flowmeter installation point where a small amount of natural gas was being diverted to another location. However, the well head take-off line flow rate was measured with an orifice meter and accounted for only a 0.5 MMSCFD. This could not have been the cause of the discrepancy in the reading.
In addition, it was determined that there was a pressure relief line branching off this line between the onshore plant and platform. The pressure relief line was tied to another line coming from another offshore platform. Initially, the pressure relief valve was believed to have been in a closed position, however, more investigation revealed that it was open. Plant operators reset the valve to its closed position and the flow rate indicated by the GC868 gradually increased over the next few hours to 18 MMSCFD. Unfortunately pressure on the line 1 increased beyond acceptable limits during this time also and operators were forced to allow the valve to re-open to satisfy maximum process pressure operating requirements.
Several days later, GE Panametrics' engineers agreed to help further investigate this process flow dilemma and brought out a complete demo GC868 system with plans to install it on the customer's 12-in. CS line coming from the other platform. Original readings were first reviewed and it was found they were had returned back to around 12 MMSCFD.
Flow readings from that lines orifice meter indicated 25.3 MMSCFD. In addition, about 0.5 MMSCFD was being diverted from the pipeline to the well take-off line upstream of the plant. Flow readings from the orifice meter on the 12-in. line from the other platform indicated 44 MMSCFD. About 1.5 MMSCFD was being diverted from the 12-in. pipeline to the well take-off line upstream of the plant. After installing the demo GC868 on the 12-in. line, it indicated 54 MMSCFD of flow coming into the plant on this line. The table below details the analysis.
Orifice Flowmeter Readings (MMSCFD)
GC868 Flow Readings (MMSCFD)
Outlet of Platform 1 (6-in. pipeline)
Well Take-Off on line from Platform 1
Outlet of Platform 2 (12-in. pipeline)
Well Take-Off on line from Platform 2
Based on the analysis above, it was discovered that there was a net flow of between 11.5-12.4 MMSCFD being by-passed from the Platform 1 line over to the Platform 2 line. The total flow to the plant was found to be the same as engineers expected, but it was coming into the on-shore processing plant in different amounts from the respective inlet pipelines. The graphic below better depicts the flow situation.
As can be seen from the data above, the total flow into the plant very closely matches the output flow accounted for by the platform orifice meters. The 1.9% difference in measured values is within the +/-2% of reading accuracy specification of the GC868 system, especially considering the accuracy limits of the orifice meters some 20 miles away. Results verify that transit-time ultrasonic technology can work very successfully in the field under ''real world'' operating conditions. The results further justify other benefits of clamp-on transducer technology including simple installation, no need to shut down the process, no wetted parts to corrode, erode or wear out, no moving parts to malfunction, and no pressure drop. All these advantages that were once though only available for liquid applications can now be realized for gas flow.
- Daryl Belock, application engineer, GE Panametrics (Waltham, MA), email@example.com
- Comments, questions: contact Dick Johnson, senior editor, firstname.lastname@example.org