Data Focus: Teradata looks East

In the Middle East, Teradata already acts for mobile companies and is active in other commercial sectors. “We are relatively new to the industry globally,” Niall O’Doherty, Teradata’s business development manager for the Emerging Industries Team, with a remit covering Europe, the Middle East and Africa, admits to Oil & Gas Middle East.

“What we’ve been trying to do is take the stories that you hear around eBay and apply those insights to oil and gas.”

The data deluge By 2020, the digital world will be 44 times as big by volume in 2009, according to a study by global market intelligence firm IDC. The upstream oil and gas industry is no different.

After all, it was the oil industry that first spurred the bulk purchase of early forms of data storage, and was at one point the largest purchaser of magnetic tape.

Other industries have caught up, with eBay announcing at the Partners Conference that they load 15,000 transactions a second through their accounting and analytics systems.

These volumes are common in oil and gas: National Oilwell Varco recently reported that rigs using new intelligent bit strings are each likely to generate data at a rate of four terabytes a day. Across the life and breadth of an average 300-400 well development project, the data volume is intimidating.

“Sensors have become very inexpensive, and so explorers are sensor-equipping everything,” Glenn Sartain, senior industry consultant – Oil & Gas at Teradata, tells Oil & Gas Middle East. “That will lead to what I call a data tsunami coming at these companies. The majority of it has been regarded as data noise, or what Chevron calls ‘data exhaust.’”

On average, 70% of an engineer or geologist’s time is spent searching for and managing data, and there are fewer experienced professionals able to make the kind of business-critical judgment calls on which successful upstream projects rely.

In light of this, oil and gas companies are recognizing that there is a lot of value in that ‘exhaust,’ but they lack in the the age of the multi-petabyte (1,000 terabytes) project, a way to navigate it.

Sartain says the digital oil field future of the industry has so far seen “a lot of hype and not a lot of substance.” Why?

Stuck in silos
Part of the problem has been that there has been undue fixation of the frequency which data is collected under the guise of ‘real-time’ digital oil field aspirations. “The idea of what is ‘real-time’ really needs to be addressed,” says O’Doherty.

“There is a cost with going down to seconds or minutes, but is there value there? If you are talking about decisions of whether to send a boat out with a spare part, do you need to know in a nanosecond?”

Teradata, taking account of the long life scales of any upstream project, aims to build business intelligence into this, so oil companies don’t install the latest expensive fiber optic everywhere just because they can.

Oilfield services giants have retained the data acquired during exploration and field development, and each has provided their own solution to specific problems in upstream projects, from G&G to wireline logging and maintenance analytics.

One result of this approach is that data accrual has been viewed by management as a cost of doing business. “The thinking has been that it has to be maintained, so you build a cost structure for it, you’re not very happy about it, but you’ll store it because you have to,” explains Sartain, a 20-year industry veteran.

“Data is actually as much of an asset as an oil well, or a pipeline, or a tanker. It needs to be mined in the same way as the reservoir that I am drilling to get at.”

The level of specialization by engineers has also exacerbated the ‘siloing’ of data stores: IOCs and NOCs didn’t call for integrated data management, so it was not developed by service companies. Scalability – a vital attribute in any business system, especially one facing a data-rich future – was ignored.

Part of the problem has been the way that operators have driven competition for service business. “We’ve driven this by having service companies that compete with each other for pieces of business,” says Sartain. “If you said to a service company that ‘we’re giving you the front-end opportunity to drill these wells but we need to have all of the information collected together in a single repository,’ they would say, ‘we can’t do that’”.

“We need to separate out the data management layer from the application and visualization layers,” says O’Doherty. “Until this is done it’s going to be a challenge to overcome some of the scalability issues the industry has.”

Additionally, just as the technological possibilities for the industry opened up in the 2000’s, rising oil prices provided operators with a disincentive to change the way they worked. “Industries make changes whenever they feel pain,” says Sartain.

who says that upstream operators have also been anaesthatised to the pain of the shortcomings in how they manage their data by ‘baked in’ contingency funding in standard business models.

Increased business risk, regulation and higher operational risk from pursuing deeper and tighter hydrocarbons in difficult environments is now allowing some of that pain to get through.

Losing time
There is also the vexed issue of non-productive time, which industry analysts believe costs the upstream sector around $8 billion a year.

The incentives to manage data better are real: knocking a day off a well-drilling timetable can save millions of dollars, and making best use of G&G data can be the difference between 20 years of oil pay or a plugged and abandoned well.

Preventing another Macondo – which official reports partly described as a failure of data communication and analysis – is an obvious and paramount priority.

“If I’m a cement vendor or a mud-logging company, I’ve got all my optimised solutions for my individual role, but the challenge is in how they are all brought together: and that’s what we’re doing,” says Sartain.

“It’s in the correlation of events that you can find the real cause. “In an environment where you were collecting data at or near real-time and were applying modeling to it, none of this would have ever occurred.”

Scalable solution
This is where Teradata comes in. Unlike other players in the market, or oilfield services firms that are turning new tricks, the company has always been about data management.

To evaluate what Teradata’s expertise in data management for other industries – particularly manufacturing – for upstream exporation and G&G, O’Doherty teamed up with Duncan Irving, a sub-surface geo-modelling and data structures specialist at Manchester University, to see if Teradata’s solutions for other industries could be applied to upstream oil and gas, from seismic to financial tracking and production stats.

Together with Irving, O’Doherty has an ambitious vision for data management and analytics for the upstream industry: the ‘Spatially-Registered Data Structure’. The SRDS is a kind of ‘Google Earth’ for seismic and g&g data, which is fully scalable, tracks historical data, and is accessible anywhere.

“Our focus now is on how you can build a sub-surface data structure within a database which is scalable up to multi-petabytes,” says O’Doherty, “one that will allow real-time analytics, that will support the digital oil field and the workflow through the upstream process, and then extend that flow of information through mid-stream to downstream.”

On the seismic side, Teradata recently acquired Aster Data, an analytic database management systems firm, which will allow Teradata customers to run pattern recognition algorithms that can perform part-automation of the time-consuming seismic analysis process, allowing fewer engineers to make better decisions more quickly.

Partners conference
The company takes the position that its current customers can sell the business much better than Teradata can itself, and organises an annual Partners User Group Conference to bring current and potential customers together to share insights and see the value of adjusting the way Teradata helps with ‘big data’.

The firm announced its fifth-generation Teradata Data Warehouse Appliance platform at the conference, which doubles the performance and triples the data capacity of its predecessor, putting Teradata at the forefront of firms capable of managing the data volumes and complexity in oil & gas.

But it’s not all about data volumes and algorithms: there’s also the human equation. In upstream, that means overcoming the knowledge squeeze as experienced engineers begin to exit the profession without sufficient mid-ranking replacement bridging the gap between them and the next generation.

O’Doherty emphasises that a change in expectations should accompany a new data management system. Ensuring the right environment for sharing data is “never to be underestimated, especially in environments where people may be coming from state-run or formerly state-run companies where there might have been differing expectations of openness, sharing, transparency, and may be reluctant to be monitored.”

ConocoPhillips
For ConocoPhillips, Teradata has been looking at knowledge management, and capturing the decision-making expertise of senior engineers in production models before they retire. This whittles down the amount of information an engineer sees so they have what they need at the right time, instead of trawling for data that an intelligent algortythmn can find.

“What [ConocoPhillips] engineers are now seeing is not what a ‘greybeard’ engineer would see; they’re seeing what a greybeard would have already pulled out from the data,” says Sartain. “So when the time comes to make a decision, you have given someone who has less experience the ability to decide based on the information that’s already been combed through.”
“If someone a year after the fact says ‘why did we decide to drill that hole there,’ today, people can’t find the Powerpoint slides, which were based on some excel spreadsheets, which were based on something else,” says O’Doherty.

“What you really want is a for all those decisions to be made in an integrated environment so that all the data that backed up that decisions is available.”

Downstream
The downstream sector has never enjoyed profit oil, and is ripe for greater efficiency savings. Western Refining, a 200,000 bpd Texas independent, reported gains in profitability and strategic planning after implementing Teradata. The company now has a single-source inventory system that updates every 15 minutes.

“Margins have always been squeezed downstream, and the fact that different operations have been traditionally run in silos adds to the complexity of how businesses run,” says Sartain. “They can now see where all their hydrocarbons are across the business. We enabled them to make real-time future decisions about what is going to be there from the supply-side.”

Western’s CIO Blake Larsen told delegates that Teradata allows the whole company to get “one version of the truth” across the business, instead of different groups – whether accounting, traders or engineers – getting different data snapshots from different times.

Tellingly, this temporarily led to frictions within the business, as some group had to change the way they worked around the new system, but now Larsen says “no-one wants to go back.”

Pitching for NOCs
Teradata is thinking as big as it gets in the Middle East, pitching to national oil companies. “We’re targeting the main players, such as Saudi Aramco and Kuwait Oil Company,” says O’Doherty.

“We’re also working on approaches towards national data repositories, something we think has been neglected by the industry of the last few years. There is a huge amount of data that is being generated and stored for governmental and regulatory purposes in different kinds of national data repositories: are they getting the full value out of that?”

“The challenge for National Oil Companies is slightly different,” says Sartain. “Their G&G side is totally siloed from their drilling and completions side, and their producing side is siloed again from drilling and completion.”

Sartain is confident that putting a unified data structure under these parts of upstream draws them closer together, leading to fewer drilling days, eventually saving “hundreds of millions of dollars.”

Aramco, says Sartain, is building real-time operational centres, has the sensors, and now needs to ask what can be achieved with the huge amounts of data they are collecting.

O’Doherty is not put off by Aramco’s famously thorough – and often time-consuming – approvals process. “Aramco has already tried to build an integrated sub-surface data store, and I think there is absolutely a requirement and a demand from Aramco.

In a way NOCs can be in a position to decide things quicker, if the right person is convinced.”

Aramco are likely to trial Teradata with “something small, something innovative, something new,” in order to prove it works, says O’Doherty. “It’s an enabler, but it’s also a competitive advantage.”

The oil and gas business is where manufacturing was ten or twelve years ago, says Sartain. “It’s following exactly the same path as industrial manufacturing, we’re just behind.”

As a result, Teradata prizes its ability to take oil and gas companies to customers in other industries: there is no reason why an airline uses in-flight data analysis of engines for predictive maintenance purposes, but a gas company that uses a similar turbine cannot.

Sartain says Teradata has been “poking and prodding” around the oil and gas industry for some time. “You get to know when an industry is ready for the capabilities that you bring to the table.”