Oilfield Technology (April 2012) – Fighting PVT Inaccuracy



The Importance of PVT and the Dangers of Inaccuracy

One of the most important sources for evaluating fluid properties and predicting reservoir performance today is PVT (Pressure, Volume, and Temperature) data.

If operators are able to generate an accurate understanding of the PVT properties of their reservoir fluids, then this crucial information will form the cornerstone of all field development decisions going forward – from reservoir simulation and recovery estimates through to production and optimization strategies. In addition, samples also need to be of PVT quality to be truly effective and accurate.

Traditionally, conventional test separators have played a crucial role in measuring PVT data with separator tests conducted to determine the changes in volumetric behavior of the reservoir fluid as it passes through the separator or separators. Samples are then reconstituted to produce a representative live sample by recombining the fluids with separator gas to match the wellhead gas/oil ratio (GOR).

However, PVT data is only effective if it is accurate. A SPE paper as far back as 1997 from Adel M Elsharkawy at Kuwait University (SPE 37441), for example, highlights the dangers of inaccurate PVT data and assumptions.

In this study, a PVT simulator was used to study the changes in reservoir gas gravity and produced gas gravity during pressure depletion in a Middle Eastern reservoir where simulation studies showed that the reservoir gas and average produced gas gravity changed by as much as 50% during the pressure depletion of the black oil reservoir.

In the absence of further PVT studies, the initial separator gas gravity was assumed to represent reservoir gas and produced gas and used to calculate crude oil and gas properties. The result was a significant underestimation of the solution gas-oil ratio and oil formation volume factor, and an overestimating of crude oil viscosity.

Using separator gas to represent both reservoir gas and average produced gas in calculating gas and oil PVT properties resulted in underestimating ultimate oil recovery by a staggering 40%.

While this is perhaps an extreme example, it’s clear that the more accurate your representation of PVT data, the more you will be able to predict reservoir behavior and ensure production optimization.

This article will examine the importance of PVT data, particularly in relation to multiphase and how subsea sampling can play such an important role in ensuring PVT and flow meter accuracy.

The Continued Growth of Flow Meters

With test lines for subsea well testing costing as much as US$60 million and the accompanying logistical challenges, the installation of subsea multiphase meters, as an alternative to well testing and as a means of optimizing recovery, have become commonplace for many operators today. Current predictions are that there are 3,300 multiphase meters installed worldwide as of 2010.

Such meters today, however, can only operate to their full potential if they are precisely calibrated and benefit from high quality volumetric sampling that reflect the changing fluid and process conditions of the reservoir. This might include an increased amount of liquid and water in the gas flow, growing water cuts, or fast changing reservoir and well characteristics.

Other factors that are also likely to result in significant variations in PVT properties might include comingled well streams from subsea tie backs; changes in water properties from sea or fresh water flooded wells; and differences in salinity between injected and reservoir water.

Furthermore, as a field starts to age, so the uncertainty of metering systems tend to increase as figure 1 illustrates.

uncertainty of metering systems

Figure 1

The Crucial Role of PVT Data

In such circumstances, effective volumetric subsea sampling and accurate PVT data and PVT quality samples can have an enormously positive effect on both the meters’ performance and field-wide production strategies.

By tracking composition changes, such as changes in fluid properties like density and viscosity (often as the reservoir is depleted), PVT data can play a key role in supporting production and fiscal allocation. Much of the inaccuracies found in all forms of allocation methods today, for example, come from operators not keeping updated PVT descriptions.

Furthermore, PVT descriptions are also vital in generating accurate reservoir models which can ultimately lead to improved reservoir recovery.

And as for meters, it is the fractional data on oil, gas, water, salinity and PVT (Pressure, Volume, and Temperature) which is crucial for calibrating multiphase meters and ensuring that they operate at their maximum potential. All multiphase meters using a gamma source must be configured with the fluid properties of oil, water and gas and ideally must reflect the changing PVT data over time.

Generating Accurate PVT Data

So how can we generate more accurate PVT data and what are the limitations of today’s current technologies?

There’s no doubt that subsea sampling, processing and the use of separators can play a key role in capturing PVT data today.

There are a number of weaknesses in using a separator, however, particularly where longer pipelines and lower pressures decrease PVT accuracy. Furthermore, using a test separator to run well tests one well at a time can also have a negative impact on the economics of the field with production having to be shut down on occasion. Figure 2 explains the steps needed to conduct a well test and provides average detail on the time when production is lost and thus profit reduced.

Stages in well-testing: Time taken/average (hours): Time affecting production (hours):
Connecting well-test equipment 2-4/3 3
Building up pressure/flow 2-4/2 1.5
Testing 2-4/2 0
Deconnecting well-test equip. 2-4/3 3
Building up pressure/flow 2-4/2 1.5
Other lost time 3 3
Total: 9-17/12 12


Figure 2

The alternative to well testing – subsea sampling, however, has also come with certain limitations in the past.

Many subsea sampling systems, for example, have previously been relatively crude, failing to generate a truly volumetric representative sample that contains fluids from all the phases.

Conventional PVT analysis can also taken weeks to be delivered to the customer with analysis often based on a limited number of samples retrieved either through wireline sampling or flow tests. While oil is relatively stable, water conductivity may change significantly during the time between the sample being taken and results received by the laboratory, therefore having a major potential impact on well stimulation operations, for example.

There is also the danger of fluid contamination where samples are sometime exposed to oil-based drillling or other reservoir production fluids.

Furthermore, subsea sampling normally takes place topside with samples taken randomly without taking note of the flow dynamics of the fluids being sampled and the original conditions in the field, such as pressures, overlooked. The result is an incomplete sample and a lack of accurate PVT data.

There are also challenges when collecting PVT data, especially in remote offshore locations, such as high pressure/high temperature environments as well as sour service fields.

A More Effective Means of Collecting PVT Data

It’s with these key drivers and challenges in mind that we have developed a more robust approach to subsea sampling and the collection of PVT data where the sample is maintained at its original pressure conditon from extraction to delivery to the surface and then transportation to the laboratory facility. Maintaining this pressure condition and the true representation of the process is crucial in providing accurate PVT analyses.

We have achieved this through an ROV (Remotely Operated Vehicle)-based subsea sampling system. While using an ROV for sporadic sampling could well be an expensive process, a complete supply sampling unit is much more cost effective.

Here’s how it works. Via the ROV, the subsea sampling system extracts and transports the sample into sampling bottles under isobaric conditions and then transports them to the surface.  Key components of the new system are an ROV operated docking sampling unit (DSU), consisting of a docking unit, a hydraulic sample extraction system and sampling bottles.

The ROV transports the sampling device from the surface vessel and docks onto a stationary subsea sampling interface (SSI) through a standard hydraulics and manipulator system. The two parts are then connected with a robust connector and barriers which are then tested to verify pressure integrity. Figure 3 illustrates the system in sampling mode, after the DSU has been docked onto the SSI.

The operation described is repeated multiple times on the same well in order to secure a number of samples over a certain time period. The result is a seamless process from sample collection to final analysis topside – from extracting a representative sample, taking to the surface, and then storing and transporting to the laboratory facility.

In this way, many of the traditional limitations of operating topside, lengthy delays, and sample contamination are addressed. Instead, measurements are taken directly from the flowline under measurable and controlled conditions with fluid properties and PVT data taken closer to real-time and closer to the wellhead.

The Rise of Virtual Flow Meters

Accurate PVT data will also play a key role as the industry sees the growth in virtual flow meters which will see, according to the research institute, NEL, the employing of software that combines distributed measurements to calculate the flow rate. For example, the pressure drop across a choke, the wellhead temperature, and the downhole pressure could be used as inputs.”

In such circumstances, PVT quality samples and PVT data will be crucial.

Accurate Sampling and Accurate PVT Data

From reservoir modelling and reservoir simulation through to comprehensive field development strategies, the accurate generation of PVT data and PVT quality samples is crucial to understanding reservoir performance today. An accurate sampling system is a highly effective means of achieving this.


View published article