Digital Energy Journal- Data quality holding real time drilling back

Join | Existing Member Login

You are Home » News » View Article

Data quality holding real time drilling back - by Jay Hollingsworth

Thursday, June 5, 2014

Sending real time drilling data to shore could be a great help in improving decision making, and ultimately lead to unmanned drilling. The biggest obstacle is currently data quality. Jay Hollingsworth, chief technical officer for Energistics, explained the current picture.

Getting real time data from drilling rigs to shore is already proving a great help in improving drilling decisions, getting more experts involved in running the well - and it provides a pathway to autonomous drilling, where there are no personnel working on risky offshore wellsites at all.

But the biggest obstacle is still the poor quality of data which operators receive, said Jay Hollingsworth, chief technological officer of Houston oil and gas standards body Energistics, speaking at the Digital Energy Journal conference in Aberdeen in February, 'doing more with drilling data'.

In fact, the mining industry appears to be beating the oil and gas industry in making fully autonomous equipment, he said. 'A few years ago the mining industry sent people to the oil industry conferences to learn how you control big industrial equipment in the field. They learned a bunch of stuff, came back a few years later, they had a movie at this workshop showing a Rio Tinto mine in Australia with no humans. Trucks drive themselves, load the ore themselves, trucks know what to do.'

'Oil industry people are disappointed we gave the mining guys some good ideas,' he said.

Better data

The biggest obstacle to using real time data from the rig to make decisions onshore is often the quality of the data, he said.

'On a drilling rig, you normally have multiple contractors, such as rig logger, LWD contractor, and mud logger. Each of these people has their own computer system on the rig, a laptop in an explosion proof box. They are gathering their own data.'

'All of this data is sent to the customer, who is supposed to do something with this information. But these different contractors on the rig are competitors and they don't play very well together.'

Sometimes companies don't even use the same clock. 'Maybe one is an hour off because one contractor doesn't realise it is daylight saving time or a time zone difference.'

Having unsynchornised clocks can be relatively easy to solve, with a software program which automatically adjusts the time from the various data feeds.

A harder problem is when companies use different readings for important measurements, for example hook load (the force of the drill string pulling down on the hook). 'Different contractors don't trust each other to measure hookload. So you'll find two or three different sensors in a row on the same piece of cable. You can't get some simple software which works out whose hook load is right.'

At the Rome SPE workshop ('Real Time Decisions While Drilling, Nov 12-14 2013), 'Italian company ENI had some slides showing how awful the real time data they get from contractors is and how hard it is to rectify everything. We spent a whole day talking about this.'

Another problem is that companies use different versions of the WITSML standard. 'The standard that most service companies support is WITMLS 1.3.1, which is 5 years old. The standard has moved on. The fact that people aren't taking up quickly newer versions of the standard is an impediment.'

Using wired drill pipe, a (fairly) new technology which enables data to be sent up the drill pipe, rather than through pulses in the mud flow, and carries faster data rates, would be a great help.
'The slowness to take up new technology is an inhibitor to data quality in this industry.'

Another difficulty with data quality is that offshore staff don't see providing good data as part of their job description.

To fix the problem, some service companies are developing tools which compare the data on the driller's morning report, with a computer calculation, he said.

A drilling data engineer in the audience commented that people are often more comfortable with what they know, even though it is unreliable and provides poor data. 'It's pretty common to go offshore and try to hook up all the systems and find people are throwing cables across the deck. Most of the operators will just listen to service companies and go their pace and that's been the way for the last couple of years.'

Case studies

One big independent US oil company was using WITSML to communicate hook load and rate of penetration data. The information was transferred into a spreadsheet in the field office, and then retyped again into a master database at company headquarters, with many opportunities for losing or mistyping data.

The company did not want data to go automatically from the drilling rig to the company database by WITSML, because it wanted to give the company onsite foreman opportunity to approve it first.
'The culture in a lot of onshore US companies is that the area foreman is king of the field office and data will move when he says,' he said.

The solution was to use a WITSML Excel adapter, so the field data is presented to the foreman as a spreadsheet. Once it is approved, it is sent onto headquarters in WITSML format. The foreman is still able to check he is happy with data, but there is no opportunity for data to get lost.

Saudi Aramco (Saudi Arabia state oil company) used WITSML to develop a connection between its wells master database and the real time drilling data, he said. 'So as they set casing, casing depth goes into the database as static information using WITSML. Nobody needs to rekey that information,' Mr Hollingsworth said. 'Saudi Aramco really believes in WITSML. They put in their drilling contracts, that if you move data from the rig to the office, it will be moved in that standard. '

'They have every imaginable service company, and every imaginable piece of software in the office. If they had 38 proprietary ways to move data from the field to the office, and 16 different applications which try to consume the data, life would be a mess.'

Pemex (Mexico state oil company) has real time operations centres in Southern Mexico run by a number of different companies, including Halliburton, Schlumberger and Petrolink. 'PEMEX wanted to give everybody a chance to prove their technology,' he said.

But for getting data from the real time operations centres to company headquarters, PEMEX mandates that the data is sent using WITSML.

Having data in a standard format in the office means that drilling engineers can see all of the data together and make faster decisions with it, rather than having a mix of different software applications which can work with different companies' proprietary systems.

Companies are not forced to use WITSML for getting data from the rig site to their operations centres (although Schlumberger and Petrolink have chosen to use WITSML for this), he said.

Statoil (Norwegian state oil company) mandates that all data from offshore to the office is in WITSML.

It also uses the same standard to move data between different software applications in the office, such as Schlumberger's Finder and Halliburton's OpenWorks.

Statoil is also keen to use Energistics' standard for reservoir data which will be released shortly called RESQML 2, and it encourages the subsurface software companies (including Halliburton, Schlumberger, Roxar and Paradigm) to all use it.

WITSML2

A new version of WITSML is under development, called WITSML2, which will have the option of sending data without tags, to reduce file sizes.

All versions of WITSML so far have been written in a programming language called 'XML' (eXtended Markup Language) which means that all of the data has a tag before and after saying what the data is. This makes it easy for computer systems to identify what the data is, and means that people can read it too.

A disadvantage is that the tags add a lot of volume to the data. A raw wireline log file could be 500 mb - converting it to XML could make it 10-20 times larger. The burden of tags can be much larger when moving really dense information such as reservoir models.

So for WITSML2, users will have the option of moving data in a new standard for untagged information, developed by the University of Illinois, called HDF5.

WITSML2 will also use a new technology called WEBSOCKET for streaming data. With WEBSOCKET, the recipient computer requests data as a stream, not as a discrete chunk (as with standard web protocols).

Another area of interest is closed loop control, or two way communication. Currently WITSML only handles data going from the rig to the shore.

With a closed loop control system, a shore computer can specify what the drilling hookload should be, and perhaps also what adjustments should be made to the hookload on a second by second basis, rather than the driller having his hand on the brake.

There will also be work on WITSML to add quality information, for example whose sensor generated the data and the accuracy of that sensor.

RESQML 2

Energistics is also launching RESQML2 later this year, a new standard for moving reservoir data between applications. Like WITSML2, it will be based on HDF5, to help reduce data file sizes.

It will enable operators to use a mix of different subsurface software applications. 'Operators want to pick and choose, they say, 'I want to use Petrel and Roxar for what it's good for, take what Roxar puts out and send to CMG for reservoir simulators.'

'CMG will support it the day it comes out, also Schlumberger and Paradigm,' he said. 'The software vendors know their customers really want to do this.'

RESQML2 won't be able to solve the problem of different subsurface software packages constructing their models in different ways. 'Petrel does pillar gridding. If you want to move a pillar grid to an application which doesn't' support pillar gridding, the standard can't help you,' he said. 'If they receive data in a gridding style which they don't support, they'll have to write software to turn it to a gridding style they do support.

Satellite communication

The problem of limits of satellite data communications is emerging as an issue with real time data, Mr Hollingsworth said.

It might be possible to reduce the data communication load by doing more processing at the rig site.

One oil major drilling data engineer in the audience said that some operators are now telling service companies they have to aggregate data at the rig site and send it back to shore together, rather than giving each service company a separate data stream.

Watch Mr Hollingsworth's talk on video and download slides at
http://www.digitalenergyjournal.com/video/1156.aspx

Associated Companies
» Energistics

CREATE A FREE MEMBERSHIP

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