You are Home   »   News   »   View Article

Making data governance sustainable

Thursday, July 24, 2014

Many oil companies find their data quality initiatives short lived or ineffective - soon after a clean-up project, the data quality starts to deteriorate. Ed Evans of NDB explained how to make the work sustainable

Oil and gas companies often have problems sustaining data quality. They have a big project to clean up their data, but then after that it starts to degrade, said Ed Evans, co-founder and managing director of UK oil and gas data and software consultancy New Digital Business (NDB), speaking at the Digital Energy Journal Stavanger conference on May 6, 'Doing more with Subsurface Data'.

This leads to a lot of rework from the DM (data management) team and another failed 'IT' project as far as the business is concerned.

The end result is a lot more short-term projects such as data clean-ups, but poorer long-term results and continued frustration for the business, he said.

Many companies are trying to tackle the problem and finding it harder than they expected. 'A common approach in oil and gas today is to say, we don't want to keep redoing things, we want to make sure that when we do the work it persists," he said.

The key to getting it done is making sure the right people take responsibility for the data, and also making sure you have an organisational structure to support it, in particular with attention given to the relationship between the DM team or IT department and the rest of the business, he said.

The industry has had a clear idea for many years of what good data management looks like, he said.

The data has a definitive source, you know what data is available within the company, you know how to find more data is you need it outside the company, and you trust the data.

Oil and gas people are often very good at solving short term problems, but not very good at looking at long term solutions, because they don't often tackle the key aspect of any project - how do we sustain a business benefit?

Conversely, if oil and gas companies decide they want to take a long term view of data, the solutions will follow. One development team said, "The asset we are working on will exist much longer than any of us working on it. They took a determined look at data management from day one, to the extent of building that into their performance contracts."


It is still typical for oil companies to have data stored all over the place, in corporate repositories and local hard-drives, mixed databases of raw data and interpreted data.

If no-one trusts it, no-one knows who is doing what with it, and no-one knows what the process is for moving data then the value of the data reduces and more time is needed in verification, Mr Evans said.

The problem is getting harder because "there are more sources of data, more data types and bigger volumes of data; on the other hand the business is demanding more from the data and from fewer experienced geoscientists and engineers,' he said.

At the end of the day, having incorrect data can lead to problems that can escalate all the way up the organisation - such as the company filing inaccurate figures for its reserves. "There's a responsibility, which goes all the way up to the CEO to make sure data used to make decisions is as accurate as can be," he said.

Data Governance Model

A good step towards improved data management is effective 'Data Governance', including defining roles and responsibilities of end-users and the DM team to achieve it.

A Data Governance Board is primarily concerned with allocating budget to data management, as appropriate for the business. The Data Governance Board also decides, how the budget is spent, and the priority of the work. "Clearly we can't tackle everything at once," he said.

The Data Governance Board has work in defining budget for on-going operations and improvement projects. It is important the Data Governance Board identify the long-term business benefit associated with any initiative even though it may be difficult to quantify.

It is only when the true benefit is understood and the project designed to deliver these results that any 'IT' or 'DM' project has any chance of success.

Putting in this sort of governance model also means that management is involved. "If data management was represented at this level, there's more authority, and assurance that whatever is decided gets done," he said.

"It helps to embed data management as a discipline within the organisation and provides the credibility to data management," he said.


"One of the things that can hold all data management initiatives back is the concept of ownership,' he said.

'People will often come reluctantly to data management meetings, but the idea that they do something about it and own the problem is often quite difficult to get over."

"A discipline chief may recognise their level of ownership for certain results (based on certain data types) but may be reluctant to get too involved in how the data is managed or the how a particular quality initiative progresses. Business engagement is critical to success but has to be managed carefully," he said.

IT and business

Often there are gaps between the IT Department/DM Team and the 'business' (the rest of the company), he said.

It can be a mistake for the business to leave the IT department to implement the technology and then don't get involved any further, because often the question of business impact can be obscured by 'IT' process.

"In the world of technical systems and data, somehow we need to overcome the resistance of the business in getting engaged with data and anything which has an IT label," he said, "and stop the IT department thinking, we're going to do it all ourselves."

For example, one technical discipline of one oil and gas company decided they wanted to upgrade Petrel software every six months and keep the last three versions of Petrel available, and expected the IT department to get on with it. The IT department take this as a mandate from the business without too much question.

"You can see some logic behind it,' Mr Evans said. 'But that's a lot of work to keep upgrading your technical system, with all the plug-ins and the integrated data. "Sometimes the implementation and maintenance [a system] has a much bigger negative impact on the business [than people expect].'

Mr Evans showed a picture of what might be called the Petrel ecosystem - all of the systems which link to a company's Petrel implementation. So every time there is a Petrel upgrade, all of the associated systems are affected.

Sometimes the business will say to the IT department, 'do what you want to do, so long as it doesn't impact the users," he said. But it is very difficult to get technology and systems implemented which don't impact the users. In fact there should be an impact - a net positive impact - on the business.

Sometimes companies rely too much on technology purchases to improve data management. "You can have a fantastic set up and the latest software but it may not improve business effectiveness," he said.

In one typical example, the IT team will install software and make it go through the various levels of testing and security analysis, fill out the forms so it can go into 'production' as planned on Thursday evening and see this as a successful implementation. Unfortunately whilst the software is available on servers the users were a long way from being able to use the software without training, data migration and a higher level of hardware.

'If anyone in the business thought, 'actually the value of implementing Petrel is not the fact you implemented it, but when you can do things better than you could with your previous software', then no-one has had that conversation with the IT project team"

"That's a really complex conversation, working out how to implement software which helps us find oil and gas better and more cost effectively than before."

The cost of software can end up being equivalent to the cost of drilling three or four wells, he said. "People say, give us the software we need, the costs are insignificant compared to drilling a well. But actually the costs can be significant these days, and the additional 'costs' in efficiency can be more important still."

IT managers are often conflicted between a desire to keep things as a steady as possible, and a desire from the business to have the latest and greatest technology. "There's no easy answer to that, it is falling between the two stools," he said.

Mind the gap

An efficient Governance Model and defined roles and responsibilities for business users and the IT/DM team can go a long way to filling in the gap between the E&P business and IT which is very common in our industry.

Business ownership of data can be made very clear, it is also important that the business owns the applications too. In order to answer the question the question 'How do we want to do Geophysics in our company'? The Discipline Lead or Function Head has to be aware of what tools are being used and how the various types of geophysical data are treated on their journey through the organisation.

"We don't need to see the data and applications as a big amorphous mass," he said.

For example you can make a list of all the software tools which are used for geophysics in the company.

You can present the list to the geophysics department and say, you decide which of these tools are important to you, and then the approach to managing data can follow from that.

Different Drivers

Ownership of data and applications is key. The IT or DM team are the facilitators for making data and software available to the end users. Any operational decisions or project initiatives must tackle the difficult question of net business impact. For example Mr Evans continued 'The completion of a Software implementation project depends on the software being adopted by the end-users rather than simply being available on a server'.

Too often the leadership in the business doesn't want to 'get involved' with technical systems and data. Clearly for the large percentage of the 'generic' IT time devoted to Oracle, SAP or Microsoft then the E&P team can afford to leave it to the IT professionals. However, business efficiency is critically dependant on having high quality technical data in most suitable technical application available when required by the business. To achieve this objective the business has to take the lead.

Associated Companies
» New Digital Business
comments powered by Disqus


To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.


Using Intelligent Character Recognition (ICR) tools on old documents to create a digital version
Piyush Pandey
from Geologix


Latest Edition January 2018
Jan 2018

Download latest and back issues


Learn more about supporting Digital Energy Journal