As computers have evolved, transient simulators for realistic field operations have become much more accessible, even in real-time. More importantly, transient simulations may even be required under steady-state conditions!
Steady state simulations of multiphase flow are widely used in the petroleum industry, partly because calculations are generally quick and require less information than the more computationally intensive transient evaluations. However, as computers have evolved, transient simulators for realistic field operations have become much more accessible, even in real-time. More importantly, transient simulations may even be required under steady-state conditions!
Correlations vs physics-based models
But let’s take a step back. The tools that are available for predicting the behavior and conditions of multiphase flow in wells and pipes may be divided into two core categories: correlations and physics-based models. Correlations are relationships between two or more physical variables, generally obtained by curve-fitting to a set of observed data (measurements). This means that correlations may only be applicable for a limited range of systems and conditions. Physics-based models are instead based on fundamental physical principles, like the conservation of mass, energy and momentum. Here, we can distinguish between steady state physics-based models and transient physics-based models.
In practice, many flow simulation tools use a combination of correlations and physics-based models. Physics-based simulation models may use correlations to describe certain phenomena, like bubble and droplet entrainment in multiphase flow or the inflow from a reservoir.
Steady state vs transient physics-based models
Steady state models are based on the assumptions that all flow conditions and properties of the system are constant with respect to time. Transient models, however, can handle conditions that change with time and assess the time-dependent impact on thermal and flow predictions. The difference between transient and steady state calculations can be illustrated using a simple example: taking a shower. Given the characteristics of the faucet, the dimensions of the hose, inlet and ambient temperatures, a steady state simulation will only provide the resulting shower water temperature and rate. A transient simulation will, in addition, tell you how long it takes before the shower temperature is pleasant, even if you make adjustments to the valves during warm-up.
Yet again, there is not a clear-cut split between the two concepts. Parts of a transient simulation tool may incorporate steady state models for certain calculations, like pumps and heat exchangers, or for certain closure relations in the set of conservation equations.
The need for transient simulations
Lower information requirements and quick calculations are central aspects to why steady state simulations of multiphase flow are widely used in the industry, but it may not offer the whole truth. It is obvious that transient simulations are vital when evaluating scenarios that are fundamentally dynamic. Typical examples are:
- System start-up and ramp-up: liquids handling and thermal behavior.
- System shutdown, including phase separation and cooldown behavior.
- Depressurization: required capacity and time, liquids handling and thermal behavior.
- Pigging operations and pigging frequency requirements, due to e.g. wax precipitation or liquid accumulation.
Furthermore, transient multiphase flow simulations may become relevant as the conditions change over time. Mature oil fields, that initially have produced single-phase oil, may experience a transition to multiphase flow in wells and pipelines as the reservoir pressure decreases, the water cut increases, or both. This means that the operation of systems that have historically been simple to analyze due to the relative simplicity of single-phase flow, may need to be analyzed using transient multiphase tools to establish an envelope for optimal operation and avoid down-time.
Less obvious, transient simulation of pipe flow may be crucial even for scenarios where the configuration of the system is considered constant. In cases with a fixed separator pressure, constant valve openings, set pump speeds, steady ambient conditions, etc. it may be tempting to apply a steady state simulation approach. However, just because the system itself is unchanging, the multiphase flow may still be unstable. A steady state simulator will generally be able to predict various flow regimes, for example the time-averaged behavior of hydrodynamic slug flow, and may even have correlations for estimation of slug statistics built in. However, such predictions are, by design, inferior to transient physics-based simulations, which can predict liquid loading and local liquid accumulation over time, terrain-induced flow instabilities, as well as other flow dynamics and thermal transients.
Transient multiphase simulations – for us the better and safer approach
Transient multiphase flow simulators, like our FLUX Simulator, are essential for the prediction of a large range of dynamic flow phenomena, from local flow instabilities to full field shutdown scenarios. For real-time predictions offered by FLUX Virtual Flow Meter or various adviser applications, a transient approach adds the benefit of taking the “history of the system” into account, for example the volume of liquids that accumulate in a wellbore or pipeline during a turndown period, or the temperature development in the casings, annuli and formation around a well. At Turbulent Flux, we combine transient multiphase flow modeling with machine learning to achieve both the highest level of accuracy and highest pace of output to help optimize operations and detect potential challenges before production levels are affected.