A cost effective solution in monitoring production with Virtual Flow Meter

INTRODUCTION

Virtual Flow Metering System (VFM) is gaining traction being a cost effective solution in monitoring production as compared to conducting well testing or having a multiphase flow meters (MPFM). VFM system is an alternative to physical, direct multiphase measurement. An Oil and Gas (O&G) production systems typically consist of a number of flowlines to an Inlet Separator via a Production Header of a processing facility. During the process, it is imperative to constantly monitor flow rates of oil, gas and water. Apart from being able to determine current well operation mode, having an understanding of the flow rates form a complete picture of the reservoirs’ production performance, which directly allows operators to make informed decision in areas of production optimisation, rate allocation, reservoir management and field prediction (McDonald & Zmeureanu, 2015).

In respect to monitoring oil and gas production, physical well testing is a conventional approach. Physical well testing requires a dedicated Test Header and Test Separator, and the configuration may result in a high capital expenditures of field development. Despites the cost impact, it is still widely being adopted, even if MPFM have been installed. It derives the fact that flow rate measurements from well test are normally adopted as a reference to calibrate multiphase flow meters  and extract information about fluid properties. On the other hand, MPFM is similarly capable of providing information about well flow rates in real time, however the use of MPFM can be expensive and requires interventions in case of any failure, significantly contributing to operational expenditure.

In contrast to well testing and MPFM, VFM may appear to be comparatively cost effective as it requires no hardware installation, estimates flow rates in real time and reflects changes of flow condition accordingly.

DESCRIPTION

There are a number of VFM methods and software being developed in the market.

The First Principle is among a method that adopts real time data from downhole measurement to make estimations of flow rates. The method appears to be most commercialised in VFM systems at time being.

VFM systems built upon the First Principle are based on a mechanistic modelling of multiphase flows in near well region, wells, pipelines and production chokes. The models are incorporated with the measurements including pressure and temperature to find accurate estimates of flow rates. An optimised algorithm is designed to adjust flow rates and other tuning parameters to minimise the mismatch between model prediction and real measurement.

Data-driven VFM is an alternative method to obtain flow measurement data with the use of integrated modelling. The approach is also being referred to as the machine learning (ML) modelling and appears to becoming popular with the advancement in ML techniques.

VFM systems with data-driven approach aims to collect field data and fitting a mathematical model to it without exact description of physical parameters of the production systems such as a wellbore, choke geometry, flowline wall thickness, etc (Bikmukhametov & Jäschke, 2020). If the model is well trained, the systems can achieve fast and highly accurate real time flow rates.

The gist of a VFM systems is to collect field data and transform it in a numerical model designed to estimate flow rates. Measurement data includes but not limited to:

  • Bottomhole pressure and temperature (PBH and TBH).
  • Wellhead pressure and temperature upstream of the choke (PWHCU and TWHCU)
  • Wellhead pressure and temperature downstream of the choke (PWHCD and TWHCD)
  • Choke opening (Cop)

ADVANTAGES

  • Retrieve real time well production rate to facilitate production optimisation
  • Less effort to compute daily production reconciliation factors
  • Ability to accurately inflow performance parameter for pressure transient analysis
  • Operation can be ensured within equipment integrity limits deriving from high quality rate estimates

References

Bikmukhametov, T., & Jäschke, J. (2020). First Principles and Machine Learning Virtual Flow Metering: A Literature Review. In Journal of Petroleum Science and Engineering (Vol. 184). Elsevier B.V. https://doi.org/10.1016/j.petrol.2019.106487

McDonald, E., & Zmeureanu, R. (2015). Development and testing of a virtual flow meter tool to monitor the performance of cooling plants. Energy Procedia, 78, 1129–1134. https://doi.org/10.1016/j.egypro.2015.11.071