NEL’s Damian Krakowiak examines the potential of virtual flow metering technology and how better testing could enable its adoption.
Validation needed. Photo from NEL.
With the variety of technological solutions for multiphase flow metering, many companies are now providing software solutions for establishing average flowrates for specific applications. Increased demand for real-time flow rate representation for both oil reservoir management and fiscal allocation are the main driving forces for the development of software-based solutions to estimate well flowrates.
Measurement of production is essential to best optimize the hydrocarbon production strategy from wells. This is achieved by performing a well test, the data from which is used to optimize the well’s production rates. Installation of physical multiphase flowmeters can be quite often problematic (calibrating, post-installation tuning, maintenance) and very expensive, however.
Virtual flow meter (VFM) systems can be an effective alternative for multiphase flowrate measurements and can be used as a backup for the existing systems. However, there is little understanding of the measurement uncertainty of VFM systems in industry. Validation of VFM systems is crucial and, although recent studies describe overall performance of virtual measurement systems, they do not evaluate the models which drive VFM systems in great depth nor over wider range of multiphase flow conditions. To provide confidence to end users, more knowledge on the performance of VFM systems is essential.
Virtual flow meter system
In the last five years, several VFM systems have been installed in most of the major producing areas, including the Gulf of Mexico, North Sea and East Asia.
VFM systems models are based on data from existing instrumentation within the system such as pressure and temperature sensors, choke valve feedback, and control valves feedback. Various mathematical models are created to ensure that together they fit and fully represent the overall system adequately. There are different approaches that can be taken to create a complete model of multiphase flow. VFM models can be created by including steady-state or transient flow simulations that are based on data reconciliation and data validation techniques. Appropriate modeling can be achieved by basing models on single components such as: reservoir, choke, valves, orifice, venture, and flowlines. Another method is based upon the complete network model that consists of many components interconnected together.
As shown in Figure 1, VFM systems can also operate in conjunction with other modeling systems, for example, well and pipework flow simulation software can be integrated with the system to facilitate real-time modeling. VFM systems can also be linked to statistical analysis methodologies, which can assist real-time trend identification.
Most of the existing virtual flow meter models are based on the principles of conservation of momentum as well as on mass and energy equations. The correct application of these equations ensures that each instrument will be also “checked” for its operational health. In addition to the measurement values from the existing instrumentation network, full data analysis and reconciliation will show live respective uncertainties for each measurement. This will be used to determine if individual instrument measurements are within uncertainty limits by changing the instrument state to alarm state.
In turn, this will help predict instrument behavior and aid decision-making on routine maintenance, highlighting any instruments that are not working or not performing within expected uncertainty.
Validation and verification
Like most measurement technologies, VFMs require period verification or calibration. On average, instruments are calibrated at yearly intervals, or when there is a significant change in process conditions or well performance. Calibration involves comparing the VFM predictions with physical flow measurements. This can normally be accomplished by performing a well test, i.e. routing the flow from the well through a test separator or a multiphase flow meter for a set period of time. If necessary, model parameters will then be adjusted to minimize the differences between the predicted and measured flow.
Research Partnership to Secure Energy for America (RPSEA) completed an evaluation into flow modeling in 2011 . The RPSEA project planned to complete a detailed study into the available VFM packages and to complete and evaluation using field data. There were three stages to the project. The first evaluated VFM systems ‘out of the box’. This meant they were not tuned before use. The second stage was to evaluate how the quantity and quality of data affects the VFM. The third and final stage was to evaluate the importance of different modelling techniques used with the VFM.
The RPSEA project originally had eight suppliers involved. It’s important to note that only seven suppliers completed stage 1, two completed stage 2, and only one completed stage 3. According to the RPSEA report, the dropout rate was so high amongst the VFM suppliers due to a lack of funding for the suppliers involved. Also, there was a lack of interest from suppliers when it became apparent that the RPSEA project could only supply simulated data, and not actual field data.
The development of VFM technology has been proceeding very quickly in recent years. However, it is generally agreed that further testing and development is required for VFM to constitute a viable replacement for physical metering. Presently, they are best used to supplement physical metering – usually in one or more of the following circumstances:
Impact on industry
Validated virtual metering systems can be extremely important in helping improve multiphase measurement. Correctly designed and tuned VFM Systems can not only support flow assurance management by ensuring successful and economical flow of hydrocarbon stream from reservoir to the end point but also greatly improve preventative maintenance.
This could extend the performance and lifecycle of all assets and also reduces safety and environmental risks, as well as operational problems, which could cost industry tens of billions of dollars every year. An accurate multiphase measurement system could potentially increase recoveries and in conjunction with tuned virtual measurement system models these recoveries may be increased.
Damian Krakowiak is a project metering engineer at NEL, a technical consultancy, R&D, testing and calibration organization across all areas of flow measurement. He has an MSc in applied instrumentation and control systems (oil & gas) from Glasgow Caledonian University.
1. Toskey, E. Research Partnership to Secure Energy for America, Evaluation of Flow Modelling – Improvements to Deepwater Subsea Measurement. Report No: 07121-1301-Task4. 2011