• Print
  • Decrease text size
  • Reset text size
  • Larger text size
08/05/2015

« Digital oil », a lever of value creation through Big Data

Oil companies are faced with growing complexity in drilling and production operations: more and more extreme weather and geographic conditions, rising costs, decreasing quality of deposits.
In order to face those technical and economic challenges, certain companies chose to invest in digital process optimization solutions. For instance, in 2002, Chevron launched the project of "digital hydrocarbon fields" named i-Field®, supported by specialists of oil infrastructures (Schlumberger, Halliburton) and of information systems (IBM, Microsoft).

i-Field®: a major long-term project

In 2002, the goal was to improve the upstream operational processes of the company (exploration, production, management of reservoirs) thanks to real-time processing of data collected on 40 strategic operational sites, spread across the globe. Those data concern drilling, reservoir status, subsurface mapping, seismicity, gravimetry and remote control of operations.

i-Field® is also based upon the deployment of thousands of captors within those 40 sites, whose size ranges from several millimetres to a few centimetres. Those captors record and transmit in real time the pressure, the temperature, the chemical or mechanical reactions and the physical properties of fluids. Thanks to the replacement of the conventional VSAT satellite technology - Very Small Aperture Terminal1 - by optical fibre and telecom networks, the data flow transiting to decisional data centres of Chevron could quickly deliver more than 1.5To/day for each equipped asset.

Finally, Chevron put in place 6 to 8 centres in the world, dedicated to data analysis and real-time decision making, which are equipped with SCADA² and relational database management systems. Each centre is conceived for one precise type of operation and allows to optimize data processing by regrouping a unit of experts and specialized analysts. Those centres fall into 3 categories:

  • Real-Time Drilling Optimization Centres ;
  • Real-Time Production Optimization Centres ;
  • Real-Time Reservoir Management.

Figure 1: 40 strategic assets in Chevron's "i-deposit"

Source: Sia Partners

All those installations rapidly allowed Chevron to follow the whole upstream production channel (Exploration, Drilling, Extraction and Reservoir) in real time.
This network of assets mainly allows to improve 3 specific key performance indicators:

  • The speed of return to full capacity after a restart or a repair thanks to preventive maintenance of installations ;
  • The Non-Production Times (NPT), a very high cost item for exploring and producing hydrocarbons. They are reduced thanks to real-time data analysis and the adaptation of technical means deployed, hunting unexpected variations in pressure gradients, which are still accountable for nearly 40% on NPT nowadays³ ;
  • The recovery factors of deposits, thanks to a closer follow-up of reservoirs and optimized asset management.

In this way, a decision centre dedicated to real-time reservoir management and to integrated asset management (RTRM/IAM)4 is divided into clearly identified business processes. Each process aims to contribute to building data analysis models that improve decision making on a specific topic and create value. Trust in predictions of reservoirs' behaviour grows if several models are used in parallel. Besides, risks are better controlled by eliminating the models that do not fit with real data. The decision-making cycle is shortened through data formatting adapted to the management process and via the use of data visualization software. On field, the injections of drilling products are optimized, or even remotely controlled, installing additional wells is facilitated and new proven reserves are registered more quickly.

Substantial benefits despite important investments

At the level of a well, the gains stemming from those installations are already important. The savings amount to up to 7 days before the starting up of a new well ($50M saved per well) ; one month in registering new P15 reserves and adjusting the DD&P6 ($12.5M saved per well) ; 2 hours on the recovery time of full production capacity after a stoppage ($3M gained after 5 restarts) ; and a reduction of the frequency of test wells before starting up, improving the test/production ratio by around 4,000 BPJ (well valuation: + $1.4M on average).

At global level, the benefits of the i-Field® project are substantial. According to IBM estimations, a $74-million investment in a mature hydrocarbon field producing a $2-billion turnover generates a 10-year ROI of 360% with a payback period of only 37 months.

Figure 2: Evolution of the ROI of the investment on 10 years

Source: Sia Partners

For Chevron, even though investments in i-Field® amounted to $5 billion during the first 5 years of the project7, then several hundreds of millions in the following years, benefits reached $700 million in 2013 and are estimated at $1 billion per year from 2016 on. Production sites show improved profitability - up to +4% for the production rate and +6% for the global recovery rate, with gains coming essentially from tertiary recovery8 - as well as a reduction of operational costs reaching 25% for a global cost of projects diminished by 2 to 4%.

An initiative taken up by other oil companies

The digital oil project launched by Chevron is thus a Big Data initiative that is very profitable in operational and financial terms. Most big oil companies currently have similar projects: "future deposit" at BP ; "smart deposit" at Shell ; "GeDIg" at Petrobras ; "digital enterprise" at Total. In total, investments in the technology of smart hydrocarbon fields reached $18.7 billion in 2011. They should climb to $33 billion by 2022, with an actuarial rate of 4.8%/year9.

Smart fields have a high cost but allow the implementation of more complex methods of assisted hydrocarbon extractions, which in turn improve deposits' profitability. Securing high-value and confidential data flows transiting between production areas and data centres has become a major challenge for those companies, in order to avoid industrial espionage or theft of information.

The success of that type of project lies in the choice of analysis models10, technologies adapted to data volumes ("big data") and teams of experts ("data scientists") set up to exploit the results at best.

Antoine Mirabel

 

Notes

[1] Very Small Aperture Terminal: satellite communication technology using dishes whose diameter is lower than 3 metres

[2] SCADA: Supervisory Control And Data Acquisition, where all captors have a unique identifier allowing to aggregate and organized collected data before classifying them in tables and relational databases

[3] Figures from a study of the Federal University of Rio de Janeiro

[4] RTRM/IAM: Real-Time Reservoir Management / Integrated Asset Management

[5] P1 reserves: these are the proven reserves. Oil companies also have P2 reserves (proven + probable) and P3 reserves (proven + probable + possible)

[6] DD&P: Depreciation, Depletion & Amortization

[7] Figures communicated by Chevron

[8] Tertiary recovery: Tertiary recovery corresponds to a category of recoverable hydrocarbon reserves of a deposit. It is the part that becomes recoverable by modifying the mobility and/or the saturation of hydrocarbons, acting mainly on its thermal or chemical properties. The global average recovery rate of deposit reserves (OOIP: Original Oil In Place) amounts to around 35% and tertiary recovery amounts to 5 to 20% of the OOIP. Oil tertiary recovery is also named EOR (Enhanced Oil Recovery)

[9] Figures from the International Energy Agency

[10] The models used aim to predict the behaviour of the subsurface reservoir and oil extraction from the subsurface to the surface. The flow of oil along the drilling towards the surface can be modelled by studying the rheological properties of oil. Phan-Tien and Tanner's model allows to calculate the fluid's constraints on the pipe wall. The calculation of those constraints is however very specific because crude oil is a non-Newtonian fluid. Krieger and Dougherty's model, which studies viscosity according to the shearing rate, allows to validate that property of non-Newtonian fluid. Within the framework of exploration or extraction operations, gravimetry or magnetic field variations allow to model the density of subsurface layers. Geostatic models help to know and map the oil concentration of a field from a limited number of drillings.

0 comment
Post a comment

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Image CAPTCHA
Enter the characters shown in the image.
Back to Top