Upstream Technology

See all articles

Democratising the science of multiphase flow modelling

A new software programme uses analytical fractional flow models to quickly and reliably simulate multiphase flow through pipelines.

Built for the digital and analytical world, the PipeFractionalFlow software has its origins in a college doctoral programme.

“It’s a science project that started in 2007,” says Anand Nagoo, chief executive of Austin, Texas-based Pipe Fractional Flow.

He wrote the code that became the foundation of his company while he was a PhD student at the University of Texas at Austin.

“It’s the first analytical multiphase flow model for fast and reliable prediction of steady state and transient multiphase flow variables along pipelines and wellbores,” he says.

PFF BUILT FOR INTEGRATION: PipeFractionalFlow was designed with workflows integration in mind and for merging with existing tools. The code can be automated with an Excel wrapper or called externally as an EXE/DLL or run using minimal computing requirements such as on a 1GB RAM Raspberry Pi for real-time Internet of Things oilfield applications.  Photo: Pipe Fractional Flow
 

Most multiphase pipe flow prediction programmes relied on correlations and empiricism originating from lab-scale conditions and fluids, he says. As a result, many of these programmes required different models for different circumstances.

“I came to the conclusion that everyone was using a different set of equations, solving things in different ways. That was problematic. That also explained why different people were getting different results,” Nagoo says.

“The problem was the experts would start with the wrong model based on a misunderstanding of fundamental multiphase flow concepts.”

His goal was to find a unified analytical framework that helped explain averaged flow behaviours in the pipeline without having to rely on the introduction of “artificial” or immeasurable variables to balance conservation equations.

PFF KEY INSIGHT: In-situ phase velocities and volume fractions (made dimensionless by fractional flow charts) are the variables that govern the transport processes of the multiphase flow, i.e. the transport of conservation quantities like total mass, momentum and energy. This is why flow patterns matter. Fractional flow charts interconnect flow patterns in smooth and continuous (analytical) ways.  Photo: Pipe Fractional Flow
 

“We looked at data in a different way,” with the goal of interconnecting and simplifying flow pattern behaviours, he says.

The software was tested against published and reproducible data sets to quantify and validate the predictability of the programme.

Along the way, he “inadvertently” gathered the world’s largest database on multiphase pipe flow using public and cross-referable data from around the world.

“We got a whole new, different understanding when we started to do things like that,” Nagoo says. “Most experts focused on one lab, their own set of problems and their own discipline.”

PFFcropAnand_Nagoo.jpg  Photo: Pipe Fractional Flow

"If you’re not modeling the leak at the local level, you don’t have any idea of how large your problem really is. "
Anand Nagoo, Pipe Fractional Flow
 

The analytical model and huge database created a “powerful way to look at problems. It gives us an understanding we could not have possibly gotten from just reading a journal or a book.”

The Pipe Fractional Flow company launched in 2016 and is “in the noisemaking phase”, he says. Its software is built with integration into existing workflows in mind.

"There’s an Excel 'wrapper' and a standalone executable that can be called from any graphical user interface," he says.

“You can use this in whatever system you always use.”

The software understands that slight inclines and declines can significantly affect pipeline flow, Nagoo says, so it can make more effective predictions in undulating horizontal wellbore, subsea flowline and hilly-terrain pipeline situations. The tool can also handle non-Newtonian flow and heavy oil flow.

“For anyone in the pipeline simulation world, these are dream features, wishes and wants,” he says.

PFF SLIGHT INCLINATIONS: In comparison to the Flow Scanner Imager production log of a Permian basin gas-condensate-water horizontal well, PipeFractionalFlow accurately reproduces the three-phase holdups and wellbore pressures along the undulating 5000-feet lateral.  Photo: Pipe Fractional Flow
 

It is possible to use the software for predicting multiphase flow along a pipeline or wellbore. It can detect leaks and blocks, monitor in-situ flow rates in real time, predict bottom hole pressure from surface data, and predict fluid volumes and pressure profiles along the pipeline.

For pipeline leak detection, he says, most methods have focused on analysing returning pressure wave data.

“The problem is that different blockages and leaks have different signatures,” Nagoo says. “The pressure profile in a pipeline changes depending on what type of leak or blockage it is.

"If it’s a long leak along a pipeline versus a short leak or sharp blockage, it gives you a different signature of pressure drops. You need a model at the local level along the pipeline that models the local change along the pipeline.”

Merely relying on tools that reflect only the magnitude of the change in pressure is not enough, he says.

“It’s the type, location and characterisation” that are important, he insists. “If you’re not modelling the leak at the local level, you don’t have any idea of how large your problem really is.”

PFF LEAK DETECTION: Elevation profile of the Middle East hilly-terrain wet gas pipeline used in the leak detection case history.  Photo: Pipe Fractional Flow

PFF LEAK MAGNITUDE: Results for different sizes of leaks at 80,000 feet from the Middle East hilly-terrain wet gas pipeline outlet.  Photo: Pipe Fractional Flow

PFF LEAK LOCATION: Results for the high rate leak at 40,000 feet and 80,000 feet from the Middle East hilly-terrain wet gas pipeline outlet.  Photo: Pipe Fractional Flow
 

The PipeFractionalFlow programme includes a component called Leak Signatures that monitors pressure in the pipeline at different stations and compares those numbers with the expected, or predicted, pressure signature.

“It’s not enough to find out you have a leak. You need to know approximately where your leak is occurring and how big the leak is,” Nagoo says.

In one hilly-terrain wet gas pipeline case history, he says, the software was used to detect the location and change of magnitude of different leaks on a 24.5-inch diameter, 21.5-mile long Middle Eastern hydrocarbon dew point pipeline, sourced from the American Gas Association (AGA) Project PR-148-110.

In simulating a high leak rate of 800 MMscfd and a low leak rate of 200 MMscfd, the programme determined the distinct leak pressure signatures that would occur at different locations in the pipeline.

Alarms would quickly sound for the low rate leak case if the observed pressures deviated by 2% at the intermediate station and by 4% at the inlet.

PFF SIMPLER IS BETTER: PipeFractionalFlow's analytical capabilities predict multiphase pipe flows without the need for numerous immeasurable tuning parameters designed to force-fit models to data. As such, the programme does not contain regressed data-fitting parameters or adjustment (“fudge”) factors.  Photo: Pipe Fractional Flow
 

Different leak signatures could be observed at stations along the pipeline when the same high-rate leak occurred 40,000 feet and 80,000 feet from outlet, confirming the programme's capability to differentiate leak locations under the same operating conditions.

Nagoo says catching small leaks before they have the potential to become big ones requires an accurate tool.

“This is a science-based tool. We don’t want to compete against other people doing multiphase flow modelling. We want to help them to do this better.”

To that end, he says, he wants to see PipeFractionalFlow available as a user-requested selection or third-party calculation engine that provides the multiphase hydrodynamics calculations for modelling software programmes such as PIPESIM, HYSYS, PROSPER, Harmony, LedaFlow, or OLGA.

“We want to make the tools themselves better in terms of speed and accuracy and capabilities,” he says.

He sees the programme as benefitting the oil industry by “democratising” a formerly complicated specialist field.

“In the past, multiphase pipe flow was a specialist field, and only a few people in the world did it well,” Nagoo says. “Now, your geologist can model multiphase pipe flows.

"All engineers in all disciplines can do this. You don’t need a specialist to do this. The engineer is your most important asset.

"We’re empowering the engineer to do the work in a faster and more reliable way with an easy-to-use tool.”

Read Next:


Laying down a track record

Saipem’s flagship pipelay vessel sailed out of the shipyard five years ago with a docket of high-profile jobs in its sights. Upstream Technology takes stock of the Castorone’s work to date, the lessons learned, and the mighty vessel’s next challenges.

15 Sep 2017
Latest news
Most read