special report doing more with subsurface, production and...

Event sponsored by:

Putting data science into perspective

Managing idle well data at PETRONAS

The IT behind subsurface data

Moving to a 'platform' approach

Helping Lundin develop a master data system

Why it’s hard to change drilling data processes

Reducing drilling costs by cutting the fax

Event Report, Doing More with Subsurface, Production and Drilling Data, Oct 5, 2015, Kuala Lumpur

Special reportDoing more

with Subsurface, Production and

Drilling DataOctober 5, 2015,

Kuala Lumpur

BETTER PLANBETTER RESULTS

BUSINESS ENGINEERING

Manage Uncertainty

CHECK

PLANACT

DO

This special edition of Digital Energy Journal is an Event Report from our forum in Kuala Lumpur on Oct 5, 2015, "Doing more with Subsurface, Production and Drilling Data".

Doing more with Subsurface, Production and Drilling Data

3

Event websitehttp://www.digitalenergyjournal.com/event/27a22.aspx

Report written by Karl Jeffery, editor of Digital Energy [email protected] Tel 44 208 150 5292

Sales manager Richard [email protected] Tel 44 208 150 5291

Conference produced by Karl Jeffery

Layout by Laura Jones, Very Vermilion Ltd

Cover art by Alexandra Mckenzie

Digital Energy Journalwww.d-e-j.com

Future Energy Publishing, 39-41 North Road, London, N7 9DP, UK www.fuenp.com

Digital Energy Journal’s conference in KualaLumpur on October 5 2015, “Doing Morewith Subsurface, Production and DrillingData”, covered a range of interesting topicsincluding clarification of data science, thechallenge of changing driller’s processes,master data at Lundin and data standards atPETRONAS.

Also how PETRONAS manages idle welldata, the IT behind subsurface data, the ad-vantages of a ‘platform’ approach to workingwith subsurface data, and how one oil com-pany reduced its drilling non-productive timesimply by asking the company to send alldrilling reports by fax.

The overall success of an oil and gas companycomes down to the decisions made by domainspecialists (geologists, geophysicists, drillingengineers) and the company senior manage-ment – so all the software and data manage-

ment systems should be geared to providingthese people with the best possible informa-tion.

But there is plenty of complexity involved inachieving this. The data systems and newtechnologies need to bring new insights to domain experts. The technical experts need tobe happy to move to using standard data for-mats, and supplying data according to stan-dard requirements.

All of the complex IT underneath subsurfacedata systems needs to work well – includingall the different software systems integratingwith each other well, and enough availabilityof subsurface IT expertise, which is very dif-ferent to standard IT expertise.

You need good structures for managing themaster data.

And perhaps most importantly, there needs tobe an acknowledgement by all that the hardesttask of all is using data analytics systems to-gether with experts’ standard workingprocesses.

Carlos Damski of Genesis Petroleum said thatwhen it comes to drilling data, it is a relativelyeasy task to gather, communicate, store andvisualise data – but actually changing people’sdrilling processes is a far more difficult task.Although it is a task which many people havesolved.

Perhaps this event report will give you someuseful ideas how to do it.

The full agenda, with links to some of the presentations and videos, is online athttp://www.digitalenergyjournal.com/event/27a22.aspx

Subsurface data in Kuala LumpurPutting data science into perspective, the challenge of changing driller’s processes, master data at Lundin anddata standards at PETRONAS – some of the topics covered at Digital Energy Journal’s conference in KualaLumpur on October 5, “Doing More with Subsurface, Production and Drilling Data”

Digital Energy Journal - Special report, Doing more with Subsurface, Production and Drilling Data, Oct 5 2015


4 Digital Energy Journal - Special report, Doing more with Subsurface, Production and Drilling Data, Oct 5 2015

Philip Lesslar – putting data science intoperspectiveTo add value, data science needs to be used together with domain expertise, said Philip Lesslar of PETRONAS,illustrating his point with a case study using data science to analyse data on fossils, in order to betterunderstand the subsurface

There is a lot of hype about data science atthe moment, but it is something which sci-entists and oil and gas companies havebeen doing for many years, said PhilipLesslar, Principal Consultant, RegulatoryCompliance and Technical Assurance,Technical Data at PETRONAS.

He was speaking at Digital Energy Journal’s conference in Kuala Lumpur onOctober 5, 2015, “Doing More with Subsurface, Production and Drilling Data”.

“The idea of data science is not new, justthe way it is being packaged.”

A common myth is that data science can re-place the need for human expertise andstructured data management, he said. “Iwant to put that myth at rest,” he said.

To illustrate this point, Mr Lesslar pre-sented a case study of an oil company proj-ect from 1987, to create a consistent andobjective way of better working with dataabout fossils found in well cuttings. Thetask required both high level knowledge offossils, and high level knowledge of datascience.

Oil and gas companies are working with anincreasing range of data types, and the vol-ume of data and the resolution of data is in-creasing.

“There is a school of thought that you canjust throw analytics at this data,” he said.“The answer is yes and no, we have to treadvery carefully.”

“Some people think we can replace all datamanagers just with data scientists. Data sci-ence is a subset of data management, notthe other way around,” he said.

Commercial hydrocarbons are getting hardto find. We have been saying that for manyyears, and it is still getting harder, he said.

To make better decisions, the businessneeds timely access to better data, and itneeds to do faster analysis of data earlier inthe value chain, he said.

Better data can also create connectionswhich trigger new ideas and concepts, hesaid.

Data science is “still an emerging field,” hesaid. “It is not well defined as an academicsubject. The interest in data science hasrisen as a result of ‘big data’”, he said.

The topic ‘data science’ can include a rangeof subjects, including data visualisation,machine learning, mathematics, statisticsand domain expertise. Of these, statisticsand mathematics are particularly important,he said.

Statistical techniques aim to help makesense out of diverse data, such as regres-sion analysis.

Mathematics can include mathematicalmethods to help you make decisions in theface of uncertainty.

Visualisation is about communicating in-formation clearly and effectively, and also

making information easier to work with, forexample showing patterns visually.

Fossil distribution

Mr Lesslar presented a case study of a proj-ect from 1987 to use data about the distri-bution of fossils to try to better understandthe subsurface.

Previously interpretations on ‘forams’ weredone by humans, and was therefore subjec-tive and not always consistent and compa-rable.

The study focussed on fossils offoraminifera, which are single celled ma-rine organisms, which can be both floaters(planktonic) and bottom dwellers (ben-thonic). They are usually about 400 mi-crons in size, and there are many differenttypes of them.

The study aimed to develop a "quantitativereference matrix" of foraminifera occur-rences for paleo environmental classifica-tion.

In other words (roughly speaking) it wouldbe a means of identifying depositionalpaleo-environments of samples fromanalysing the foraminifera in them, basedon other samples which had a similar mixof foraminifera.

Because the number of species are so large,it is very hard to do with human interpreta-tion. But the computer analysis can calcu-late clusters - groups of foraminifera whichare more likely to be present in certain for-mations.

Benthonic foraminifera are sensitive to envi-ronmental conditions, or in other words theirpresence or absence will reflect the “paleo en-vironment”, ie the condition that piece of rockhas been in, over geological time.

Mr Lesslar did an analysis of rock samplesfrom 250 wells in Malaysian waters and the

5



South China Sea. This included 100 sam-ples from each well, and finding between20 and 250 different species in each sam-ple, average of 120.

This multiplies to about 3m species occurences altogether.

The data was analysed using cluster analy-sis, which is “one of the more importanttechniques in the whole array of what datascientists do,” he said.

Cluster analysis enables us to visualise ‘n-dimensional’ data using a ‘dendogram’or tree diagram. The clustering looks at‘measures of proximity’, how close onegroup of samples is to another group.

From a geological point of view, rock sam-ples with a similar mix of foraminifera probably have a similar history in how theywere formed.

The larger the data set is, the more sure youcan be about the patterns.

“That's the power of doing a cluster analy-sis,” he said.

“You end up being able to make probabilis-tic calculations, saying that for a certainpaleo-environment, the probability of a cer-tain species being present is x.”

“Every time there's an occurrence you cal-culate the positive probability of it beingpresent or absent,” he said.

The final “identification matrix” covered411 different species and 13 depositionalenvironments with a probability for each(ie a probability of species no 200 being inenvironment no 7).

By assimilating and statistically organisingthe total set of sample data, the softwarecan then generate a ‘probabilistic matrix’,

ie say that a certain environment wouldhave a certain probability of certain speciesbeing present.

“A probabilistic computer assisted interpre-tation system that would remove the incon-sistency associated with humaninterpretation was developed.”

So for a new well, if you can take samplesand see what ‘forams’ are in it, the com-puter can tell you which formations thatwell is likely to have been drilled through.

This could all help the seismic interpretersin understanding what they are looking at.

Watch Philip’s talk on video and downloadslides athttp://www.digitalenergyjournal.com/video/1636.aspx

6

Managing idle well data at PETRONASPETRONAS embarked on a project to keep better track of its idle wells, and make sure it had plans in place toimprove them



Malaysian state oil company PETRONAS re-cently embarked on a major project to improvehow it handled data about idle wells, and madesure it had plans in place to reactivate them.

Karenjit Kaur Narinjan Singh Executive Petro-leum Engineering Data, Technical Data withPETRONAS, explained how the system worked.

She was speaking at Digital Energy Journal’sconference in Kuala Lumpur on October 5, 2015,“Doing More with Subsurface, Production andDrilling Data”.

An idle well is simply defined as a well which isnot producing anything, but is not yet abandoned.

PETRONAS defines it specifically as a “stringthat has not produced or injected continuouslyfor 90 days.”

Some other parts of the world use a different def-inition, such as a well which has not producedcontinuously for 6 months over the past 6 years,Ms Singh said.

The ‘idleness’ can be due to operational reasons,such as sand in the well, failure in artificial liftequipment, or a delay waiting for additional per-forations to be made in the well.

Or it could be idle due to reservoir reasons, suchas no oil and gas is flowing into the well, or thewell is producing at too high gas oil ratio.

PETRONAS sets a target of well ‘availability’ ornon-idleness, she said, in order to achieve its pro-duction targets.

PETRONAS senior management want to be able

to better prioritise decisions about which idlewells should be given maintenance or workovers,she said. The maintenance or workover workshould be higher priority if it means the biggestincrease in production for the least cost.

PETRONAS senior management also want toget a much better idea about which wells are be-coming idle.

Ms Singh’s team built a software system whichwould gather data about idle wells from the fieldin a standard format, and present the data to man-agement, so they could better make decisions.

Inconsistent

Before the project was started, PETRONAS hada problem of no standard system within the com-pany for classifying idle wells, with inconsistentinformation flow.

Data was being managed by Excel spreadsheetsand not always being shared, or with a numberof different versions of the spreadsheet in circu-lation.

When the company management wanted dataabout idle wells, it would take some time for staffto provide it, because checks needed to be madeabout the accuracy of the information.

The engineers were required to submit monthlyreports about idle wells in their region, and thistypically required one week’s work to gather thedata.

Sometimes staff already had a well activation

plan in place, and senior managers were notaware of it.

All of this meant that management don’t get areal overview of the well.

New system

The new idle well management system connectswith PETRONAS’ production database, whichhas information about production data across thewhole company.

It will auto generate an “idle candidate,” in otherwords it will generate alerts for production engi-neers saying ‘it looks like this well is idle, canyou verify whether it is or not’.

If the well is idle, the production engineers areasked to provide a reason.

The idle well reporting has been incorporated inthe production engineers’ existing productiondata management software, so they don’t have tolearn a new system.

Company management have a dashboard whichwill show them where the idle wells are, and howlong they haven’t been producing. They can alsosee which wells are becoming idle and whichhave been idle for a while.

They can see how close they are to the target, andtry to fill the gaps.

Management can see which regions have highpotential for more production, and put their focuson those areas.

The dashboard has visual indicators, showing foreach region the total number of idle wells, howmuch this number is growing (more idle wells)and how many wells have been restored, andhow many have a reactivation plan, whether theplan for the previous month was achieved, andhow much additional production would begained for the re-activation or restoration plan.

The reactivation plan usually starts with a tech-nical study. If an idle well does not have a tech-nical study within 2 months of becoming idle, analert is sent to management.

Download Ms Karenjit’s slides at http://www.digitalenergyjournal.com/event/27a22.aspx

7Digital Energy Journal - Special report, Doing more with Subsurface, Production and Drilling Data, Oct 5 2015


Tony Joseph – the IT work behind subsurfacedataSubsurface IT managers and specialists do a critical task in helping the subsurface team identify projects.Tony Joseph, Geoscience Analyst with Murphy Oil, explained how it works

Subsurface IT is very different from usual IT,and might be better in a separate departmentwithin the geology and geophysics (G+G) busi-ness, suggested Tony Joseph, Geoscience Ana-lyst with Murphy Oil.

He was speaking at Digital Energy Journal’sconference in Kuala Lumpur on October 5,2015, “Doing More with Subsurface, Produc-tion and Drilling Data”.

Mr Joseph has been involved with GIS and geo-science software for 16 years, including work-ing with Shell, PETRONAS and Hess, and hisremarks were about his experience at oil com-panies in general.

The IT department is typically concerned withissues like Microsoft Office installations in thecompany, ERP, networks and servers, he said.

But the subsurface IT department is concernedwith making sure that the company can do asmuch as possible with the data it has available,to better find oil and gas.

This includes critical tasks like making sure thedata fits together, spotting errors, and choosingsoftware packages – and supporting the geolo-gists and geophysicists do their best work.

For example, consider that the subsurface IT

manager is probably the best person to knowwhich subsurface software is the fastest. He cansee that doing one task takes 3 minutes on onesoftware package, and 10 minutes on another.This is very useful information to know.

Too often, software purchasing decisions aremade by senior management, who are mainlyconcerned with how much the software costs,and are persuaded by a fancy presentation fromthe software company sales person, he said.

Subsurface software is rarely properly testedand evaluated in the company by qualified soft-ware specialists before it is implemented, hesaid.

Companies are not usually very interested ingetting a full understanding of what technologyis available and how much it can help them.

Actually, all of the subsurface software pack-ages “have their strengths and weaknesses,” hesaid. “Not one single software is perfect.”

One complexity is that the different subsurfacesoftware systems all use different symbols fordifferent types of wells, and different colourcoding.

There are also many advances in IT which canmake subsurface work much faster, includingcloud computing, faster hard drives, fibre opticcommunications and blade servers, but they arenot used as much as they could be.

A desirable end goal for oil and gas companiesis to have all of the subsurface data integratedtogether on a geographical information system(GIS), including visualising the petroleum sys-tem, showing all of the seismic data, and wellsdata, he said.

The subsurface IT department can have manydifferent job titles. Mr Joseph’s job title is “Geo-science Analyst”, and some companies use“Geoscience Technician”. But his role is basi-cally to work with the subsurface data.

Workflow

Mr Joseph described a standard workflow toidentify oil and gas prospects, from the perspec-

tive of a subsurface IT technician.

The first IT task is to put together a base map,gathering together all available geological in-formation about the region, including knownwells, known oilfields, water depths and facili-ties. There may be ‘georeference tags’ labellingpossible geological features.

Well locations on the maps should match withwell locations on the well logs and final well re-ports, and can also be checked against publicdata about well locations (from companies suchas IHS).

The geotechnician will also do the task of con-verting seismic data from time to depth, basedon the velocity model and depth conversion for-mulae.

Meanwhile, the geophysicists will do seismicinterpretation work, including marking out thehorizons (dividing line between formation lay-ers) and faults. The faults need to be identified,marked with polygons (multisided shapes) andclassified as to what kind of fault they are.

A petroleum systems expert and geologist canthen try to work out where the oil and gas mightbe, looking for the four elements of reservoir,source, seal and trap.

Sometimes this will involve creating a ‘com-mon risk segment’ map, working out the likeli-hood of the presence of each of the riskelements of the petroleum system, reservoir,source, seal and charge.

Once prospects have been identified, the geo-science technician will classify and map them,making an estimation of the volume of the stor-age area and amount of oil and gas which mightbe in them.

Finally this data is presented to management tomake a final decision about where to drill, in an“evaluation montage,” he said.

Download Tony’s slides athttp://www.digitalenergyjournal.com/event/27a22.aspx



PETROSYS – helping Lundin develop masterdataPetrosys helped Lundin Malaysia develop a master data management system. Rob Bruinsma of Petrosysexplained how it worked

Oil and gas subsurface data management companyPetrosys recently ran a project with Lundin Petro-leum Malaysia, to help the company develop a mas-ter data management system.

“Better data management will benefit everybody,”said Rob Bruinsma, Senior Database Analyst / Ad-ministrator with Petrosys.

He was speaking at Digital Energy Journal’s con-ference in Kuala Lumpur on October 5, 2015,“Doing More with Subsurface, Production andDrilling Data”.

“It's pretty clear that communication and trust be-tween business partners will improve if data is man-aged better than it was before.”

Lundin wanted to implement the system so it couldimprove the data quality, security, accessibility andtrustworthiness of its data. It would also help Lundinto comply with the regulatory data reporting re-quirements, which, in Malaysia, are set byPETRONAS.

In particular, Lundin wanted to centralise data, tomake the same data available to all projects and re-duce the time that people spend on finding the rightdata.

Benefits

Before implementing the system, Lundin had datastored on many different software systems and diskdrives, making it hard to find.

“One of the first things people would like to see im-proved is the time to find data,” he said.

“And when you find data, you want to be sure of itsquality, and you want to be sure it is the same dataother people are using in the company,” he said. “It’sall difficult if it’s not centralised,” he said.

Other advantages of the centralised data were im-proved data security, and reduced risk of error.

Also, mistakes are more likely to be identified.

“When many people see the same data over time,many people will be able to correct it, say 'this isn'tquite right can you please change it here,” he said.

If people in the company like data, they will use thedata more and more over time, he said.

You want to reach a point where people say, “I'mnot interested in going to my C drive there's a worldwhere people say, my data in the central database isof such good quality I’m going to go there first,” hesaid.

System chosen

The project team selected a Linux based server forthe data management system, which had good per-formance and was inexpensive. It chose Oracle forthe data management system (because that is theplatform on which Petrosys delivers the richestfunctionality).

Lundin chose the PPDM standard database struc-ture, developed by the Professional Petroleum DataManagement Association.

It was keen to use a non-proprietary database struc-ture, ie one developed by an association rather thanby a commercial company. This will mean that avariety of software applications from different com-panies would be able to run from the database, andit would not be tied down to just working with onesoftware company.

“Petrosys believes that PPDM is one of the betteropen databases on offer,” Mr Bruinsma said. “Wehave been associated with PPDM for 15 years ormore.”

Before data loading

“Before we do the data loading it’s very importantto construct what the rules are for this data,” he said.For example, what labelling structure you are goingto use.

You don’t want duplicate data. This means you haveto define specifically what ‘duplicate data’ is. Defin-ing this, and creating other business rules, is themost important starting point.

If you don’t do these things, your database will be“thousands of records of meaningless drivel no-oneknows how to deal with,” he said.

You also need to decide what you are going to store.

For example you can store data about formationtops (what depth of the well different formations

start). But some of this data might also be includedas part of the well header, where people may alsobe likely to look for it.

Having detailed data can be helpful, for example inmaking errors stand out and become more obvious,but you need to balance the need for detail with theneed for having a data management task which youcan complete and sustain.

Probably there will be some duplication in the enddatabase, since there is no perfect system.

Another issue is how you will work with unstruc-tured data, including documents and videos.

For these, you might just put a note in the masterdatabase that the data exists and where the locationis on a networked drive. These manual links can beadded over time, rather than all at the beginning. Inthe mean time you will be dependent on automateddocument catalogue and search systems as well.

You might want to limit the sort of documentswhich can be included.

Data loading

Once you have made a decision to go to a masterdata management environment, the initial data load-ing should be done as quickly as possible, Mr Bru-insma recommends. Decisions can be takenafterwards about what improvements need to bemade, and they can be prioritised.

Data is initially loaded in bulk into the database, in-cluding Kingdom data and spreadsheet data.

Data can be loaded into intermediate ‘staging areas’before being added to the master database.

It is also good to have automatic processes. “Youdon't want to be manually keeping track of what isloaded,” he said. Although automated data checkscould make it much slower to load in data, he said.

Once loaded, data should be given an automatedquality score that is based on business rules for thedata in question; while a confidence score shouldbe attributed to the same data, by users who edit thisdata. Data quality/confidence monitoring tools area very important aspect of data management - todrive data quality improvement over time.

Download Rob’s talk on video and slides athttp://www.digitalenergyjournal.com/video/1717.aspx

9Digital Energy Journal - Special report, Doing more with Subsurface, Production and Drilling Data, Oct 5 2015


Carlos Damski – how can we change drillingprocesses?Data gathering, processing and visualisation is comparatively easy, said Carlos Damski of Genesis Petroleum.It is changing companies’ processes which is hard

When companies are trying to reduce costs, theyusually look in 3 areas: people, technology andprocess, said Carlos Damski of Genesis Petro-leum.

He was speaking at Digital Energy Journal’sconference in Kuala Lumpur on October 5,2015, “Doing More with Subsurface, Produc-tion and Drilling Data”.

Cutting people is the first thing companies do,but there is a limit to how much it can be done,he said.

When they look to cut technology costs, for ex-ample by looking for products with lowerprices, there is also a limit, because you can’toperate without technology, he said.

But companies tend not to look very hard at re-ducing the costs of their processes.

As an example of how much money can besaved by processes, consider how budget air-lines have managed to remove a large amountof costs from operating airlines, although theyspend the same on staff (pilots) and technology(the aeroplane) as full price carriers.

Improving processes in oil and gas basicallymeans maximising operational efficiency, orgetting more out of the existing assets.

Data is key to achieving this. “If you want totransform or improve our business we have tolook harder and harder in the data, and the data’srole in the business,” he said.

Mr Damski sees data management like a pyra-mid. At the base you have all the work to ac-quire and store data. At this level, everythingonly has cost, for software, hardware and con-nectivity.

The value from the data comes higher up thepyramid when you start working with it. Ulti-mately you can do something like provide use-ful information to a drilling engineer who canwork out how to reduce time and cost of drillingoperations.

Most parts are easy

Most elements of the data management processare fairly easy.

“No-one says, ‘I have a problem with data ac-quisition,’” he said. “This is very easy.”

Data storage is cheap and easy. There have beenproblems (like people storing data on their ownhard drives) – but this can be easily solved (forexample by disabling people’s ability to storedata on their own hard drives).

Doing quality control on data is not a difficulttask, with many tools available to help you iden-tify and remove bad data.

Providing access to data, including from multi-ple data sources, is well understood.

Visualising data is also fairly easy. For example,doing statistical techniques on data to work outthe biggest sources of non-productive time,drawing pie charts of your production over thepast 6 months.

Processes

But where it gets really difficult, or where com-panies usually get stuck, is in using the data toimprove their processes. “The world of drillingand the world of computing have to come to-gether,” he said.

Improving processes requires a continuous im-provement feedback loop, which is independentof the people and personality issues.

You need to try to incorporate people’s experi-ence better into the process.It would be very helpful to store lessons learnedfor the next generation.

We may have mastered how to store data, butwe do not yet have technology which can storesomeone’s expertise in a computer system. Fornow, we are reliant on human expertise.

“It’s pretty much a personal thing in terms ofimproving processes,” he said. “We store les-sons learned, and try to systemise more what weare doing.”

Mr Damski also demonstrated the concepts withsome business cases.

Download Carlos’ talk on video and slides athttp://www.digitalenergyjournal.com/video/1721.aspx



Jess Kozman - reducing drilling costs byavoiding faxOne South American company forced all its drilling rigs to submit data by spreadsheet rather than fax,leading to big savings, said Jess Kozman, Regional Representative of PPDM

Drilling downtime is a huge part of total upstreamoperational costs, said Jess Kozman, Asia Pacificrepresentative of PPDM (the Professional Petro-leum Data Management association).

He was speaking at Digital Energy Journal’s con-ference in Kuala Lumpur on October 5, 2015,“Doing More with Subsurface, Production andDrilling Data”.

Total drilling downtime cost to the industry can becalculated as three times the actual rental cost of therig, so if the rental is $600,000 a day, downtime costcould be as much as $1.8m a day.

Many companies are interested in using ‘big data’techniques on drilling data, for example using somekind of predictive analytics to try to work out whichcomponents will fail.

But companies typically collect and store a lot ofdata but don’t do anything with it, he said. It is easyto analyse data and generate a bunch of numbers,but it doesn’t necessarily mean they are useful inbusiness, he said.

For data to be useful in oil and gas decision making,we often need to combine ‘managed data’, situa-tional awareness data, analytics data and perhapshistorical data together.

This makes it a much more complicated challengethan in other fields of analytics, such as internetsearch, where (for example) the age of the data isoften not a concern.

“The idea of bringing all the data together in oneplace is not really new,” he said. But now, comput-ers are trying to bring the data together in a com-puter system, rather than a person’s head.

As an example, consider the idea of ‘play based ex-ploration’, where an oil company looks for oil byconsidering where all the factors leading to oil mightbe present (source, charge, trap and seal).

50 years ago, this task would have been done byone expert geologist, gathering all of the data in hishead. Today, people are trying to do this by com-puter.

Some data is more valuable than others. For exam-ple, someone who has been involved in productionfrom a certain basin over a long period of timemight be able to tell you that the most valuable datawas collected from a certain era of drilling, for ex-ample when the first appraisal wells were beingdrilled.

But a data manager, who is aiming to set up a com-puterised data management system from all the doc-uments and reports gathered from field operations(some of which may be handwritten) may decide toeither start with the most recent data and work back-wards, or start with the oldest exploration data andwork forward. With both of these approaches themost valuable data might be entered into the systemlast.

NPT case study

Mr Kozman presented a case study of an oil and gascompany in South America, which wanted to re-duce the non-productive time for 50 land rigs oper-ating in a single region, using 6 months ofperformance data.

The company chose to limit the scope of the project(to 6 months of data and one region) so that theamount of data involved wouldn’t be overwhelm-ing.

Out of the 50 rigs, 38 were providing data byspreadsheet, and 12 were providing data by fax(which was then re-typed in company headquar-ters).

The spreadsheet data included notes about rig ac-tivity (when the rig was drilling, when it was circu-lating mud, when it was waiting for supplies), allentered in a different category on the spreadsheet.

So there were a number of different categories of‘non-productive time’.

The oil company found that the 12 rigs sending databy fax were reporting an average of 65 days of non-productive time over 6 months, whilst the 38 rigsusing spreadsheets reported an average of 15 days.

Most of the additional nonproductive time was dueto delays in transmitting critical data to the head-quarters office, where the final budgetary decisionsfor maintenance are approved.

One thing the data management project clearlyshowed was the importance of asking the 12 rig op-erators using faxes to start using spreadsheets, sodata between the rigs could be properly compared.Eliminating the data re-entry at company headquar-ters also reduced the time needed to make decisionsabout maintenance and rig logistics.

No investment was required for this data manage-ment project, apart from buying Microsoft Exceland providing training, communications and changemanagement.

If the rigs using faxes could show non-productivetime similar to the other rigs once they were usingspreadsheets, that could be calculated as a saving of$23m over 6 months, Mr Kozman said.

Also, by no longer using faxes, the company wasable to free up a large amount of time of expert per-sonnel who don’t have to type in data from faxes,he said.

“We spend a lot of time saying 'spreadsheets are nota viable long term database store. But in this case, itwas improvement over a faxed pdf. I’ll take win-nings where I can get them.”

The company would ultimately like to get a fullyautomated rig reporting system, or database systemswhich would give more structure, checks and rulesto the data entry process. But this would take manymore months to implement, he said.

Download Jess’ slides at http://www.digitalenergyjournal.com/event/27a22.aspx

HAS YOUR TRAINING BUDGET

BEEN CUT?

Just a click away from expanding your potential.

Designed to replicate the high standards users receive from our face-to-face classroom-based training, Petrosys Online Training will provide a self-paced and convenient format that allows you to log in wherever, whenever suits your workload.

Courses offer:• In-depth theory and tutorials.• Full auditability – allowing you to track your progress.• Expert technical audio guidance.• Lower costs – eliminating travel and expenses.• Get started immediately – simply click and buy!

INTRODUCTORY OFFERS AVAILABLE NOWwww.petrosys.com.au/training/online-courses

Petrosys Online Training is a cost effective way to continue developing your skill set and knowledge.

D1042

Low Cost. High Value.

List of attendees 'Doing more with Subsurface Production and Drilling Data, Oct 5, 2015

What did you enjoy most about the event?

“ “All the speakersdelivered a veryclear message.It's like being ina classroom.

Content of the presenters“ Excellent

speakers andrefreshments. “Q&A session

and Dr Carlos& Jess Kozman'spresentation “networking

Melissa Yap, Consultant, AccentureSean Chin, Consultant, AccentureChristin Jong, Consultant, AccentureNur Nadiatul Hidayah, HSEMS and DCMSManager, ARECATECH SDN BHDAhmad Syalaby Muhamad Shamsuddin,Operation Manager, ARECATECH SDN BHDMohd Ridwan Sulaiman, Head of Geophysic, Asian Geos Sdn BhdAgus Hadijanto, Remote Operations Services Account Manager, Baker HughesLalitha Selvarasan, Beicip Technology Solutions Sdn BhdGrecia Echenique, Business DevelopmentManager, BeicipTecsolSamantha Perng, Head of PRISM (Production System & Management) Sup,Carigali Hess Operating CompanyLam Yin Hoong, Analyst, Production System Support, Carigali Hess OperatingCompanyLiau Ooi Keng, Analyst, Production SystemSupport, Carigali Hess Operating CompanyNur Atiah Binti Azhar, Technical Consultancy Manager, Cekap TechnicalServices Sdn BhdFatin Syafiqah Nur binti Abdul Halim, Project Engineer, Cekap Technical ServicesSdn BhdKarl Jeffery, Editor, Digital Energy JournalZakir Mat Darus, Lead Data Technician,EnQuestSiti Marlina M Ishan, Subsurface TechnicalSupport, EQ Petroleum ProductionBadrulisham Baharudin, SubSurface Technical Wellwork Operations, ExxonMobil Malaysia

Richard Mcintyre, Finding PetroleumCarlos Damski, Managing Director, GenesisPetroleum TechnologiesHassan Sabirin, Managing Director, Geo-ZenFirdaus Abdul Karim, Operation Manager,IHSZaiton Mohd Izham Ng, Data Analyst, IHSNur Afinni Ismail, Data Analyst, IHSMalaysia Sdn BhdAlexandra Noeb, Independent Data ServicesMohd Ismawira Bin Mohd Ismail, Executive Director, Integrasi Erat Sdn BhdSyed Mohamad Daniel Syed Idris, BusinessManager, Integrasi Erat Sdn BhdDismas Bismo, Technical Sales Asia Pa-cific, Kongsberg Oil and Gas TechnologiesJohn Chapman, Asia Pacific Technical SalesManager, LandmarkNur Khalidah Khalid, Subsurface DataManagement Assistant, Lundin MalaysiaDeano Maling, Subsurface Data Manager,Lundin PetroleumTony Joseph, Murphy OilDaniel Chan, Program Director, NetAppKarenjit Kaur Narinjan Singh, Data Executive, PCSBAsma Hanim Ahmad, PETRONASNoor Fadhilah Mohd Raes, Data StandardsManagement, PETRONASKing Chai Ngu, Manager, PETRONASNur Fatin Bt Mustafa, PETRONASChew Wei Liang, PETRONASDoreen Oo, PETRONASNoor Fadhilah Mohd Raes, Data StandardsManagement, PETRONAS

Samsurin Welch, Head, Digital Strategy,PETRONASZulkifli Taha, Head, Regulatory Compliance & Technical Assurance, Technical Data, Technical Global,PETRONASDaljit Singh Sukhdev Singh, Manager,Petronas Carigali Sdn BhdPhilip Lesslar, Manager, Technical Assurance - Data, PETRONAS E&PAnu Venugopal, Office Administrator, PetrosysRob Bruinsma, Senior Database Analyst /Administrator, PetrosysJess Kozman, Manager - SE Asia, PPDMSiti Zubaidah Abu Bakar, Data Management, RepsolSathiyaseelan Sathiyamoorthi, Team LeadSubsurface Technology, Repsol MalaysiaNathelly Wee, PreSales Systems Engineer,SAS InsituteLina Tong, Principal Solutions Consultant,SAS Institute Sdn BhdKok Yew Lim, Solutions Architect, SASMalaysiaNIK SAUFI BIN NIK MURAD, StrategyAnalyst, Searie Group Of CompanyMohd Hairul Nizam Jappiry, Manager-Solutions, Sedia Teguh Sdn BhdSteven Yew, Business Development Manager, SenergySteven Yew, Business Development Manager, SenergyChin-Fah Heoh, Managing Consultant,Storage Networking AcademyDuncan Irving, Oil and Gas Practice Lead(EMEA/APJ), Teradata



� � ��

� � ��

� �

� ��

� � ��

� �

� � ��

� � � � � �

� � ��

��)",��-)��

�

��

��

��

�� !�

�� "! ��

�� !�� !�� #$�� %&��

'��(�� )��*��

��+��)",��-)��

��583 Orchard Road

#06-01 Forum Singapore 238884

www.cltechconsulting.com

+65 9220 7396 [email protected]

As the industry re-aligns to “lower for longer” commodity pricing, the most successful operators are re-tasking data management professionals to work with operational and engineering data types. Making sure they have the industry best practice skills to enable lower operating costs and reduced risk by using all of your facilities data. Schedule a facilitated workshop with CLTech Consulting today.

special report doing more with subsurface, production and...

Documents