Business Forum: September 5, 1996 (Houston) and September 17, 1996 (Nottingham)
Agenda and Notes.
Migration SIG: September 6, 1996 (Houston) and September 18, 1996 (Nottingham)
Agenda and Notes.
Note: These notes are subject to review. Members are welcome to send comments to Alan Doniger at Doniger@POSC.org. Some presentation slide files are linked into the notes. Additional links to presentations will be added as they are supplied. Notable changes to the meeting notes will be summarized here.
Version 1. October 24. First published version.
These meeting notes supplement the meeting materials distributed to attendees. Issues and action items are highlighted. Sessions are arranged according to the meeting agenda.
Questions about POSC in general or about these meeting notes can be directed to Alan Doniger (Doniger@POSC.org) or Paul Maton (Maton@POSC.org).
Bill Bartz introduced Bob Pindell of Texaco as the newly elected POSC Board Chairman. He thanked Thore Langeland of Statoil for his service as chairman this past year. He announced that Mousa Gaddourah of Saudi Aramco was elected as the Board vice-chairman succeeding Bob Pindell. He congratulated Jean-Paul Marbeau of CGG Petrosystems on his re-election as At-Large Director for a two-year term.
Bob and Jean-Paul reviewed the topics discussed at the September 3rd POSC Board meeting. These topics included: finance, election of board members and officers, annual planning session (October 22-24), POSC/CAESAR, and a report from the POSC/PPDM Committee.
Bertrand reviewed the history of joint work between POSC and PPDM, the formation of a POSC Board Committee, and the recent meetings between the committee and representatives of the PPDM Board. He described these meetings as being useful and said that contacts will continue until the committee and PPDM agree on proposals for future collaboration and/or consolidation.
See Bertrand's presentation in a textual form.
Questions and Answers:
- Bertrand was asked to explain how Project Discovery relates to the committee's work. See Bertrand's extended response.
- Bertrand characterized the probability of success for Project Discovery with the expression "cautiously optimistic".
Bill described the request being made to members to supply feedback on the draft POSC mission and ENDS statements for 1997. Attendees received a handout of the page posted on the POSC Web Site that contains the text of the proposal. There is provision on the Web page for members to send feedback comments directly to Bill Bartz via email. Feedback may also be sent to any member of the POSC Board of Directors.
Bill explained a technical change which resulted from consultation with POSC's lawyers on the subject of neutrality. References to standards have been replaced with references to specifications when referring to POSC's deliverables. To clarify the relationship between specifications and standards, the phrase "specifications for use as standards" has been edited into the POSC mission statement.
Bill emphasized that members will determine the content of the next release of POSC specifications. He said that "needs" rather than "nice-to-have's" must be identified. Bill noted the sense of urgency and the need to achieve "return on investment" in POSC by members.
Questions and Answers:
- An additional ENDS statement was recommended for consideration by Frank Coe (Chevron) stating the need for balance between "functionality" and "stability" as the specifications evolve.
- Hamish Wilson (Paras) asked whether the predominant position of the demand side in POSC governance might change as implementation proceeds. Bill noted the changes leading to the presence of three supplier at-large directors on the Board.
E1. ISO / STEP Introduction: Mission, Participating Industries, and Benefits; Richard Wandmacher, ISO TC 184 SC4 Chair in Houston; Sheila Lewis, ISO TC 184 SC4 Quality Committee - Process Improvement Team Leader in Nottingham
Alan Doniger introduced the session, explaining the history of contacts between POSC and ISO / STEP, especially since the start of the POSC/CAESAR project in 1995. Dick and Sheila explained the vision and mission of the ISO organization that includes STEP. They reviewed the history of STEP work, the industries involved, and the ISO standards process. Sheila noted the challenge that SC4 faces to move from addressing only product data to addressing all industrial data. Finally, they characterized some areas in which ISO and POSC could cooperate. These included POSC representing the oil and gas industry in international standards work, collaboration on business and technical aspects of standards, and collaboration in liaisons with other standards efforts.
The presentation (in PowerPoint).
Questions and Answers:
- Dick estimated the cost of a single STEP Application Protocol through the standards process to be about $1.5 million.
- Dick defined "interoperability" in STEP as meaning consistency in common portions of different Application Protocols. He explained that the need for this kind of interoperability is being questioned.
Svein Heglund (in Houston) and Nils Sandsmark (in Nottingham) added a word about the close working relationship of POSC/CAESAR and certain STEP-related activities, such as EPISTLE. They announced that an agreement was reached last week with the sponsors of AP221 to share reference values ("class libraries") with POSC/CAESAR and for POSC/CAESAR to administer the maintenance of these values.
Dick Wandmacher concluded by saying that the STEP architecture and methodology was defined at least six years ago and that it is time to examine the way STEP operates to address current requirements.
F1 Plans to Merge POSC / CAESAR with POSC; David Archer and Svein Heglund in Houston; David Archer and Johan Skutle in Nottingham
David described the work-in-progress to achieve the merger of the POSC/CAESAR work into POSC. He reviewed the points presented in June that indicate the direction being followed, i.e. no forced merger of data models, the completion of each work effort, continued collaboration to resolve differences, and the marketing of the benefits of Take-Up.
David said that the agreement is being renegotiated to extend the project through 1997 and for POSC to assume maintenance at the beginning of 1998. He reported that the POSC/CAESAR implementation team wants to adopt as much POSC infrastructure and tools as possible, e.g. DAEF, exchange operations, data model browser, etc. The investigation of how POSC/CAESAR requirements may be addressed in future versions of Epicentre continues.
Snapshot B is available and is being reviewed. POSC/CAESAR has agreed to perform the administration of the class library of STEP's AP221.
Regarding personnel, Svein Heglund is returning to Saga to be replaced by Johan Skutle as Project Manager. Tore Aven is returning to Norsk Hydro to be replaced by two current staff members, Magne Valen-Sendstad (class library group) and Geir Kildal (implementation group).
Svein Heglund and Johan Skutle reviewed the current work plan noting that a combined Snapshot C&D will be completed at year-end and that two major production releases are planned in 1997. They indicated that the product model would consider extensions in areas such as drilling, finance, shipping, and subsea facilities. They outlined plans to begin to pursue ISO standardization with an intent to issue a working draft in the autumn of 1997. Regarding Take-Up projects, they spoke about near-term decisions by Norwegian offshore projects to use the POSC/CAESAR work in setting up facility data warehouses. They emphasized the need to increase the number of participating software suppliers, engineering and construction companies, and equipment and system suppliers. They concluded by noting the importance of marketing the benefits of using the results to the industry. The 1997 budget estimate is 18 million NOK with contributions from oil companies (five at 2 million), engineering companies (four at 1 million), and government (4 million).
David concluded the Houston session by expression appreciation for Svein's contribution to the project and by wishing him success in his new job.
Questions and Answers:
- Steven Daum asked David Archer when clearer merger plans would be available. David replied that POSC and POSC/CAESAR will do more work in common, but that there will be no push to merge in the short term.
- Tom Gulman asked Johan Skutle whether the scope of POSC/CAESAR's work is the same as ISO/STEP Application Protocol (AP) 221. Johan explained that the scope of POSC/CAESAR is larger than AP 221.
- Michael Ring noted that most of the data model work has already been done. What remains is to complete the standard values, i.e. class libraries, from the current 70% complete status.
Nottingham ACTION ITEM F1:POSC and POSC/CAESAR should write an issue paper addressing plans / guidelines for "using" (extending, deploying, seeking de jure standardization, etc.) Epicentre and the POSC/CAESAR product model in the near and middle term, i.e. pending future convergence.
Bill described the survey that has been conducted by Frank Walsh. Mr. Walsh conducted one-hour interviews with about forty asset managers chosen from a list of two hundred. Mr. Walsh started by asking each executive what he/she values most and went on to ask for opinions about POSC. In general, Mr. Walsh found three equal groups of executives: (1) those interested primarily in cost cutting and bottom-line results, (2) those who are primarily motivated by technology, and (3) those seeking to implement knowledge work.
The next step is to form two focus groups of about ten people each. The focus groups will be asked to respond to what Mr. Walsh heard in the interviews. It is hoped that a final report will be ready before the annual planning meeting, October 22-24.
Houston ACTION ITEM G1:Members are asked to recommend asset managers for the focus groups by sending their names and contact information to Marise Mikulis (POSC Houston) by September 13.
Bill mentioned other sensing activities, including participation in the Quiet Revolution with Cambridge Energy Research Associates and ongoing direct contacts with suppliers, oil companies, standards groups, and research organizations. All of these activities are intended to help ensure that POSC's work is going in the right direction.
Tom described the recently defined IT strategy adopted by Mobil. The strategy definition process began with interviews of 52 business people. Business units were involved in implementing the strategy.
Single vendors were chosen for two aspects of the E&P computing environment: (1) master data stores and (2) integrated environment (for applications). Tom concluded by describing a benchmarking process which involved interviewing other oil companies to compare IT strategies with the Mobil strategy
Tom's presentation (in Power Point).
Nottingham Questions and Answers:
- Clay Harter (Chevron) asked what would happen if a business unit needs something special. Tom replied that exceptions will require higher support charges and that R&D would be outside of the IT strategy.
- Nico de Rooij (Shell) asked Tom if he would make the benchmarking questions available. Tom agreed to do so.
- Aage Haldorsen (Statoil) asked how Mobil planned to make the transition to POSC-based products from current products which are closer to PPDM. Tom referred to Project Discovery and said that Mobil will move to products as they become based on the results of Project Discovery. Mobil is counting on the success of Project Discovery. Mobil is on the Project Discovery Steering Committee.
- Tom was asked if Mobil would be looking for new applications with new functionality. He replied that added functionality may become a requirement over time as Mobil re-engineers business processes.
G3 Host Executive Presentation; Les Amidei, Texaco (Houston), Peter Cook, British Geological Society (Nottingham)
In Houston, Les Amidei, Texaco's ITD General Manager, described recent changes in the IT business drivers and how the focus of IT has changed for Texaco. Les explained the transition period that Texaco is in between "build" and "buy". Based on the longevity of both assets and data as compared with the relatively short life of application products, Texaco has concentrated on building asset data stores. This is illustrated by the well-known Kern River project success story.
Houston Questions and Answers:
- When asked why 75% of IT projects "fail", Les explained that one or more of the goals of working fast, producing quality, and staying within budget are often missed. He recognized that E&P is a high-risk business and accepted that E&P IT projects should be allowed to take on some degree of risk.
- When asked how IT strategy is linked with business strategy, Les explained that IT was re-organized around "customers" with relationship specialists to translate back and forth.
- Bill Bartz asked what Texaco is doing to help manage change. Les observed that resistance to change is mostly subconscious and passive. Adequate training and development must be provided. The biggest change is learning to think "customer-centric".
In Nottingham, Peter Cook presented a thorough overview of the activities of BGS -- past, present, and future plans. He described both the national and international roles of BGS in mapping, monitoring, advising, and R&D. He characterized the core programs as including geophysical surveys, hydrocarbon surveys, geophysical monitoring, geochemical surveys, geotechnical surveys, etc. BGS has a staff of 770 and a 37 million GBP annual budget of which 35% comes from commercial contracts, 20% from government contracts, and 45% from the government science budget. Dr. Cook described the wealth of data that BGS holds going back 160 years and including over 200 public data bases. Onshore E&P data goes back 100 years. Offshore data goes back 30 years. He expressed eagerness for a partnership with POSC. He referred to a recently approved "Field Base" collaboration with the UK DTI, ARK Geophysics and PECC, to European Union sponsored work (including Norway), and to the pending Geological Electronic Information Exchange System (GEIXS) involving 15 countries and Matra Datavision. For the future, Dr. Cook warned oil companies that they are living on their past. He referred to a recent UK DTI policy not to maintain data but to have BGS do so.
Opening Statements:
- Caesar Offshore Project / Svein Heglund in Houston and Johan Skutle in Nottingham
Scope: Surface Facility Design and Operation
Deliverable: specifications (product model) and implementation technology
Progress: Half-way; start in 1995, end projected for end of 1997 (then folded into POSC)
Organization: Full-time, funded work team, Steering Committee of partners, 300 companies queried for requirements and to give feedback
Project Administration: By Saga Petroleum
Take-Up: Working toward commitments in parallel with specification development
Leverage: POSC work and STEP work
POSC Role: joint contracted project participant; will maintain and enhance after POSC/CAESAR project ends
- Project Discovery / Frank Coe in Houston
Scope: Seismic Acquisition
Deliverable: Subsetting and Projection feedback to POSC; relational projection of Epicentre suitable for PPDM adoption for delivery through POSC and PPDM (as migration from V3.x) and with acceptable costs.
Progress: first phase mostly complete; first deliverable this year; completion in 1998
Organization: Steering Committee, open participation, no contract, core working group of three people
POSC Role: one of ten participating organizations.
- GEMSIG / Bill Quinlivan in Houston
Scope: GeoShare Data Model scope
Status: Started over two years ago; currently on hold
Deliverable: Common Business Object definitions and mappings to GeoShare Data Model and Epicentre
Organization: Voluntary participation; no contract
POSC Role: Positive contribution
- LightSIP / Alain Perrot in Houston and Gerard Huard in Nottingham
Scope: Data Access and Exchange Facility (DAEF) product development; applies to full scope of E&P life cycle
Objective: to promote the development and use of new POSC-compliant applications
Deliverable: DAEF product and promotion of its use
Status: Developer selected; beginning of development phase; project duration is 14 months
Organization: Formal collaboration of oil company sponsors (Elf, Shell, Statoil); contract with development company (IBM) and subcontractor (Prism Technologies); Project Management Group consists of the operating oil company and suppliers; Design Authority Team consists of all participants; official start around September 10; results in 14 months
POSC Role: Supplied specifications and sample implementation; proved suitability of the specifications and good advice on technology and strategy
Alain provided the following slides as a reference.
- RESCUE / Mark Dobin in Houston and Knut Kirkemo in Nottingham
Scope: reservoir characterization process; 3D reservoir models through upscaling as input to reservoir simulation
Objective: to develop prototypes to demonstrate interoperability using a POSC data store
Leverage: prior work of SAVE
Status: started earlier this year; suppliers working on common representation; halfway through overall; phase 2 work for '97 under discussion; demonstrated interoperability in the ILAB's
Project Administration: hired SAIC
Organization: contract among the participants (difficult to find mutually acceptable wording); steering committee; six sponsors; three vendors
POSC Role: consulting; mapping to Epicentre; appreciation expressed to POSC for its support
- SAVE / Todd Little in Houston and Mike Wells in Nottingham
Scope: reservoir simulation life cycle, especially simulation and feeder data
Objective: improved Epicentre footprint
Status: started two years ago; active on and off since; currently in the process of restarting; more testing scheduled; still along way from deployment (the real goal); steering committee met September 16
Organization: Seven oil companies and nine vendors participated; contract; major shuffle of vendors through acquisitions; funding model being re-evaluated; buyer pull from oil companies ambivalent; sales job needed to get the anticipated applications
Leverage: interface to RESCUE
Deliverables: validation of Epicentre; testing of applications running on DAEF products from UniSQL and POSC (sample implementation)
Project Administration: hired SAIC (has had positive and negative aspects); lack of domain expertise is a concern
POSC Role: technical support (with appreciation expressed to Robert Aydelotte and Iain Morison)
- Open Spirit / Clay Harter in Houston and Ben Weltevrede in Nottingham
Scope: geological and geophysical interpretation initially
Objective: plug-and-play, multi-vendor application environment; legacy transition to POSC; incremental, efficient application development
Status: a new alliance in the formative stage; first core team meeting on September 20
Deliverables: commercialized an object application framework focused on application integration and smooth transition from legacy systems; development and run-time product licensing
Success Factors: oil company support, supplier Take-up, POSC support (e.g. from current Interoperability and Business Objects initiative), to find a neutral custodian with needed skills
Organization: patterned after LightSIP with sponsors and a custodian company; Request for Bids for custodian role will be issued in the fall.
POSC Role: supplier of new "interoperability" specifications; invited to participate
- URGENT / Aage Haldorsen in Houston and Bill Martin in Nottingham
Scope: full E&P life cycle
Objective: to determine how business processes that span the life cycle can adapt to advanced well and other new technology
Deliverables: business process model; business case studies analyzing the impact of E&P and IT business practice changes; to define requirements for new applications and services
Status: officially started in February '96
Organization: Agip (study of Adriatic gas field), Statoil (study of open area offshore and small producing oil field with exploration in the area), and Norsk Hydro (study of advanced well usage in producing field); EU funding 50%; partnership consortium agreement; Partners Board; work done by oil companies and consultancies
Project Administration: work assistance and coordination from PDS, Paras, and The Information Store
POSC Role: non-voting Board member; support for proposed IT requirements
- WIME / Rolf Rolfsen in Houston and Nigel Goodwin in Nottingham
Scope: well design, drilling, completion, operation, workover, plug-and-abandon
Objective: to validate Epicentre; to define drilling activity "codes"
Leverage: related to both subsurface (Epicentre) and surface facility (POSC/CAESAR) work
Deliverables: standard activity coding
Status: started one year ago; work to start in the fall
Organization: Statoil and Saga are committed; BP has participated; Statoil is funding the initiation stage; open workshops planned at first; a larger project may follow; to use RFC to define scope for larger project
Project Administration: management by Statoil; coordination by Essence
POSC Role: POSC and POSC/CAESAR staff participating; concerned about POSC area expertise and resource availability
Houston Questions and Answers:
- Bill Bartz asked about lessons learned. Svein: learned to be user managed and to do detailed planning. Frank: learned that cycle time to update Epicentre is too long (about 18 months). Bill Q. learned to be end-user driven and to leverage other work. He suggested that POSC put together an alliance "starter kit". Alain learned that it is difficult to design a business model that is favorable for all parties. Mark learned to address a common 3D representation at two levels: export format and C++ library to break-up the learning curve. Todd learned that a good start needs strong buyer pull and a willing market and that it is more difficult than expected to keep momentum going and achieve deployed commercial applications. Todd added that changes in vendor companies from acquisitions has been difficult to overcome and that the two oil companies that were ready to implement went through large re-organizations. Todd also learned that experts from POSC, such as Robert and Iain, are necessary.
- David Zeh described the CGM*PIP Alliance which operated in 1991 and 1992. He solicited financial participation from 45 oil companies and suppliers. Sixteen companies joined the effort, ensuring adequate funding. The goal was focused and realizable. Phase one was completed in nine months. Phase two was completed a year later. David acted as project manager. He believes that having a paid staff helped.
- Laramie Winczewski asked what could be done to ensure coordination among alliances. Alain identified POSC's role. Bill Q. urged open publication of results. Frank encouraged feedback be fed into the POSC change process. Todd also pointed to POSC's role in coordination.
- Michael Ring asked what could POSC do to help each alliance. Svein asked for additional POSC data model staff. Frank asked for a reduction in the release cycle time. Todd asked for expert POSC help, such as that received earlier by SAVE. Rolf asked POSC to help identify the right contacts within member organizations. Clay asked POSC to sponsor workshops to help start alliances and advised alliances not to depend on volunteer work. Aage asked for additional POSC staff members in Europe.
- James Gazewood (BLM) suggested that alliance teams do concentrated work for four to six weeks at a time. Paul Maton asked if there are other effective work management models. Aage observed that the initial sponsors should do some work to get an alliance started. Svein endorsed the need for a full-time work team.
Houston ACTION ITEM H1:Summarize panel/member suggestions for POSC actions to help alliances.
Nottingham Questions and Answers:
- Hamish Wilson (Paras) asked whether these alliances were chosen based on collective business needs or based on the individual interests of the founding participants. Mike Wells replied that the choices depended on a combination of individual interests and perceived industry needs (by oil company visionaries). Ben Weltevrede explained that Spirit was an in-house development which came to be inconsistent with Shell's buy-not-build policy and so is being offered through the Open Spirit Alliance to other oil companies. Nigel Goodwin explained that Statoil's commitment was essential to the formation of WIME. Gerard Huard explained that LightSIP was first an Elf initiative to motivate industrial development. Later, Elf looked for partners to help sponsor the project.
- Hamish Wilson asked rhetorically whether POSC should analyze the coverage of the current alliances and seek to fill the gaps. Stewart Robinson (UK DTI) responded that it would be enough to address and deliver solutions to current problems. He noted that the same alliances existed a year ago. Aage Haldorsen (Statoil) pointed out that URGENT is intending to find out where to focus 'alliance-type' work from a business point of view. Knut Kirkemo (Statoil) pointed to the need for more oil company commitment and outspokenness if alliances are to make more progress. Stewart Robinson cautioned that it is not known what products are needed. He observed that each alliance was started due to the work of one strong person. Knut Kirkemo referred again to URGENT as being needed to determine oil company requirements.
- Michael Ring asked whether there are enough people to support all of the alliances or whether the same small group of people are trying to support all of them. Mike Wells replied that there are not enough people. He illustrated this with an experience in SAVE in which needed oil company resources could not be made available. Knut Kirkemo asked whether it would be possible to assign different portions of the life cycle to different oil companies so no single oil company would have to support "all" alliances.
- Nico de Rooij (Shell) spoke in favor of alliances that are based on contractual commitments from participating companies rather than those based on voluntary participation. Pierre Chevalier (Cap) added that more sponsors are needed to ensure that alliances can produce deliverables. Aage Haldorsen (Statoil) stated that there can be no single model for all alliances. As for WIME, it is sometimes necessary for a single sponsor company to get things started before additional sponsors can be added.
- David Roper (Sysdrill) characterized the alliances in three groups: those intended to produce commercial products (e.g. LightSIP and Open Spirit), those defining future requirements and specifications (e.g. POSC/CAESAR and URGENT), and the others -- which, he said, should not exist as alliances. POSC should take responsibility to validate and adequately test the specifications. Independent alliances shouldn't have to do this work. David Archer accepted this point as having some validity, but stated that alliances, such as SAVE, have also worked out important problems outside from the validation of POSC specifications. Mike Wells agreed noting that validation of POSC specifications was one of the goals of SAVE. Another goal was to improve integration with or without the use of POSC specifications.
- Tim Haynes (The Information Store) asked what changed in the SAVE alliance over the past 18 months. Mike Wells replied that there have been many company take-overs and that it has been difficult to restart the project after a pause in the work. He noted that oil companies seem to have de-emphasized POSC work.
Art reviewed the member workshops on compliance and the status of the follow-on activities, especially progress on a prototype Web-based product registry and updated Compliance Statement Templates.
Houston Questions and Answers:
- Josina Arfman (TriTeal) suggested that the POSC registry have open access and be linked with The Open Group's Product Registry. Art replied that coordination with The Open Group is ongoing.
Nottingham Questions and Answers:
- Dan Schenck asked if there are plans to help suppliers determine which profiles are close to the compliance coverage of one of their products. Art replied that this could be done.
- Glen Kendall (Oilfield Systems) asked if problems are anticipated for products that don't exactly fit any defined profile. Art explained that the intention is for profiles to be defined based on compliance patterns of actual products.
- Hamish Wilson (Paras) asked who will run the registry service. Art replied that POSC is producing the prototype by end-of-year '96. No decision has been made on who will offer the registry service for the long term.
- Age Haldorsen (Statoil) asked how deep the compliance components will go in the data model area. Art characterized the intention to define data model compliance components that correspond to widely used subsets. No specific compliance components have been defined yet.
- Bertrand du Castel (Schlumberger) asked why POSC is in the compliance business when this work can be done by private companies. Alan Doniger explained that the registry work did not involve examining products to determine their degree of compliance. That would be done by product suppliers, possibly with help from private compliance verification services. POSC's work is just to manage a repository of product compliance information provided by product suppliers. Once the prototype is ready, additional clarification will be needed to define the roles of suppliers, buyers, and POSC.
- Pierre Chevalier (Cap Gemini) asked whether dynamic compliance testing is being addressed. Art replied that this is outside the current scope of activities requested by members in the recent workshops.
- Jean-Paul Marbeau asked who would define profiles. Art replied that POSC will coordinate the definition of the compliance component "framework". Profiles will be derived by observing patterns of compliance of actual products.
Houston Suggestions:
- Executive Understanding Survey Report.
- Oil Company Experiences.
Nottingham Suggestions:
- Compliance
- Alliance formation / driving and POSC's role
- Project DISCOVERY (if at a milestone)
- Update from the Board POSC/PPDM Committee
- ISO/STEP: EXPRESS V2 (?)
- PSTI Report on its Alliances
- National Data Repositories by Stewart Robinson (for BF)
The Business Case Report will be posted on the POSC Web Site before the end of September.
Houston ACTION ITEM G1:Members are asked to recommend asset managers for the focus groups by sending their names and contact information to Marise Mikulis (POSC Houston).
Houston ACTION ITEM H1:Summarize panel/member suggestions for POSC actions to help alliances.
Nottingham ACTION ITEM F1:POSC and POSC/CAESAR should write an issue paper addressing plans / guidelines for "using" (extending, deploying, seeking de jure standardization, etc.) Epicentre and the POSC/CAESAR product model in the near and middle term, i.e. pending future convergence
Issue: Will oil companies release a larger data set to POSC?
Jan described the expanded scope of activities covered by Snapshot B. There are over 4500 (references) classes now. There were only a few additions to the data model. These were in the areas of activity, class, specific and typical, information content, and association. He estimated that would be between 30 and 40 thousand (reference) classes once all of the work has been completed. He estimated that there would be approximately 7000 in Snapshot C/D.
Houston Questions and Answers:
- Alain Perrot asked why the Epicentre terminology for "reference values" is not used by POSC/CAESAR. Jan explained that the "class library" terminology is used in the STEP community.
Nottingham Questions and Answers:
- Jan was questioned about the work ahead to define "geometry". Jan indicated that previous work on geometry by ISO/STEP and/or by POSC on Epicentre could be used. Jan was asked whether POSC/CAESAR was required to use ISO/STEP geometry as defined in the STEP Integrated Resources in order to be compliant. It was pointed out that Epicentre geometry differed from STEP geometry because Epicentre addressed subsurface geometry from a geoscience point of view, e.g. STEP geometry does not address non-manifold geometry. Jan concluded saying that POSC/CAESAR is trying to capture an overall structure from both STEP and POSC work. It was pointed out that there are other international standardization efforts that may be relevant on this subject, including the Open GIS Consortium (OGC), Project OMEGA (an EU funded project involving Matra Datavision and BRGM). [Editor's note: Other related work: (1) There is a newly formed special interest group on GIS in the OMG. (2) POSC has been actively involved in the ISO/SQL Multi-Media Spatial standardization process for several years.]
- Jan clarified that Snapshot C/D will not be backward compatible with Snapshot B and that no migration assistance has been offered to anyone trying to implement directly from the snapshots. Such assistance is planned for the formal releases.
- David Roper observed that it is difficult to determine the taxonomy of a "part" from the class library definition. Jan replied that the "industry" does not have a well-defined "part" taxonomy and that POSC/CAESAR could not force one to be defined. For example, the definition of a "pump" is not definitive.
Jenny reviewed the results of the joint work to evaluate POSC/CAESAR requirements in an Epicentre context as discussed in workshops earlier this year. She went on to characterize the requirements in three groups based on expected level of impact. Short-term items are of small magnitude will be proposed for consideration along with other proposed changes in the next release. Medium-term items have greater impact but fit within the current Epicentre structure. These items may not be addressable in the next release. The most difficult items may have significant impact. Further study is needed on these items to search for new and better approaches and solutions.
Houston Questions and Answers:
- Frank Coe asked when the first production release of POSC/CAESAR is scheduled and was told that it will be in the first quarter of 1997.
- Alain Perrot asked if the introduction of high-level business objects would help address some of the more difficult POSC/CAESAR requirements. Jenny acknowledged that this was a possibility and that it would be considered as the work proceeds. Jan Sullivan observed that the POSC/CAESAR data model contained concepts that were at a higher level than some in Epicentre.
- Alain Marechal observed that the STEP community is having great difficulty in achieving interoperability across diverse Application Protocols in and between industries and that this may suggest the need for more than one data model to exist.
Nottingham Questions and Answers:
- Bertrand de Castel (Schlumberger) observed that this examination of the POSC/CAESAR business requirements in light of prior POSC work is what would have been done first at the start of the POSC/CAESAR project. Jan Sullivan explained that Snapshot A represented an expression of the POSC/CAESAR requirements. Bertrand asked why the POSC/CAESAR requirements were being examined again now. Jenny replied that the current work is focused on finding solutions.
- Jenny explained that some POSC/CAESAR requirements can be satisfied in an Epicentre context without solving any of the difficult problems. She added that the solutions to the difficult problems need not be the same as the solutions reached by the POSC/CAESAR team.
- Hamish Wilson asked if there are more fundamental projects for the POSC resources to be addressing, whether these are current work efforts, such as PPDM coordination, or new work efforts, such as financial system coordination. Michael Ring pointed out that POSC and POSC/CAESAR have the "well" in common and that this is a strong reason to continue to work toward eventual convergence. Barrie Wells (Robertson) added that POSC and POSC/CAESAR are addressing two parts of the same problem. He suggested that POSC should not take a short-term view. Jenny added that the amount of effort being expended on POSC/CAESAR requirements analysis is very small. She noted that in about one year the work of POSC/CAESAR will become the responsibility of POSC to maintain and that there are very significant financial savings possible as a result of the application of the POSC/CAESAR work. David Roper concluded this discussion by referring to the POSC Mission statement which states "throughout the asset life cycle".
Jenny's presentation (in Power Point).
Cary described the content of the Version 2.1 release CD-ROM's. He explained the expanded use of the Browser and the presence of complete data sets as well as sample implementations. He advised users of the CD-ROM's to check the POSC Web site for errata which will be posted as errors are identified.
Houston Questions and Answers:
- In response to questions about the continued use of FrameReader for the specifications and Browser, Cary explained that POSC would like to publish in another form, such as HTML or Acrobat.
Nottingham Questions and Answers:
- There was a similar discussion to Houston about technical issues related to publishing documents in HTML or Acrobat instead of FrameReader format.
Cary and Jenny's presentation (in Power Point).
Dan explained the process which has begun to define the scope and content of the next release. He offered the suggestion that the next release would be significant enough to be called Version 3.0, but he cautioned that the version number would be designated based on the accepted scope and content. He covered some of the specifications, aside from data model, data access, and data exchange, that may need updates in the next release, e.g. the Base Computer Standards.
Dan announced a member workshop to discuss elements that may be addressed in the next release on September 19 in Nottingham. He told the attendees that a similar workshop is being planned for mid-October in Houston. [Editor's note: A Houston workshop was held on October 15.]
Dan listed some of the possible sources of proposed content for the next release and reminded members that their input and feedback is fundamental to the process of identifying and prioritizing these items. The list included property RFT feedback, problem tracker items, preferred values, alliance feedback, POSC/CAESAR requirements, POSC/CAESAR class libraries, GIS standards efforts, document management requirements, exchange operations enhancements, data access improvements, graphic metafile enhancements, and BCS update.
Dan anticipated a provision Work Plan document would be ready by the end of the year. He spoke about the possibility of doing more member testing as part of the release development process.
Houston Questions and Answers:
- Frank Coe (Chevron) asked whether there could be an interim release to address high priority issues. Dan replied that this was possible depending on what the member input and feedback indicates.
- Tom Nash (Mobil) asked what the "next release" workshops were intended to accomplish and what kind of people should attend. Dan explained that the purpose is to obtain member input as to the "needs" and to get feedback on the "needs" expressed by others, i.e. what are your "three top" needs for make the specification work (better) for you?" John Bloom (Chevron) added that the basic decision to have smaller interim or larger major release next should be resolved as soon as possible.
- In response to a question, Dan reminded members that there is an email list (property-rft@posc.org) for open discussion about the current "property" RFT.
Houston ACTION ITEM E2: POSC will set up a new email discussion list (next-release@posc.org) to support the process of defining the scope and content of the next release.
Nottingham Questions and Answers:
- In response to a question about the Open GIS Consortium (OGC), Doug Benson (Oracle) stated that OGC addresses the use of GIS in many industries - even US federal forestry management.
- Interest in member testing was indicated by several attendees. The need for an early test plan was observed.
- Bill Bartz pointed out the need for members to identify "needs" rather than "nice-to-have's" in support of strong, pragmatic Take-Up activity. Nico de Rooij pointed out the need for balance between smaller, quicker releases to get needed changes out and larger, longer releases to avoid the cost of repeated migration.
In Houston, Bill Quinlivan began the session by reviewing the results of the two workshops held on this subject: in February (Houston) and in July (London). The July workshop consisted of ten presentations which were summarized by Bill and the other two presenters. Bill summarized his presentation on Business Objects. Art Culbertson summarized three presentation on defining fine-grained objects over Epicentre and supporting an object-oriented data access technique. Clay Harter summarized six presentations that focused on aspects of defining high-level objects in an application integration-enabling framework. Clay concluded the session with a review of the concluding discussion from the July workshop.
In Nottingham, Alan Doniger reviewed the results of the third workshop held earlier that week in Nottingham. This workshop was focused on refining proposals for actual work efforts. Two break-out groups discussed aspects of the proposed work. One group looked at defining business objects and relating them to Epicentre. The other group looked at defining an application architecture (including business objects) to enable true application interoperability. The results of the workshop will be used as the basis for updating the provisional Work Plan document and for moving this process forward. Attendees were encouraged to participate in this work through the "interop" email discussion list and the POSC Web Site. Nico de Rooij added some details about the discussions of the first break-out group related to the need for tools to help people navigate through business object definitions, i.e. a semantic definition browser.
Houston Questions and Answers:
- Several members spoke in support of Bill's recommendation to have end-users involved in defining business objects.
- John Bloom (Chevron) asked how "common" business objects can be if they are closely tied to business processes which are varied and which evolve over time. Bill Q. replied that business objects are defined and valid only for a specified set of business activities. Alan Doniger (POSC) pointed out that the term "common" business objects as used by the OMG, for example, means business objects that are valid across "all" industries.
- Several members supported the overall effort and encouraged further work on this subject.
The combined Houston presentation (in Power Point).
In Houston, Paul Ford started the session by explaining how the current effort evolved from shared network services among some of the Alaska operators. Cat concluded the session by describing the data coverage and usage of the InfoBank, including exchange (EDI), shared data (e.g. well header), common data (with multiple owners), and independent data (with separate owners). The Request for Information indicated that the InfoBank should be built based on standards, be scaleable, be bought-not-built, use consulting expertise, use a prime contractor concept, use multiple sources of funding, be well-defined, have a long-term commercial commitment, and be phased.
In Nottingham, Paul Maton gave the same presentation given in Houston by Paul Ford and Cat Larrea.
Paul's presentation (in Power Point).
Cat's presentation (in Power Point) and session notes (not available).
In Houston, there were four Podium speakers:
- Alain Perrot (Elf) gave an update on the status of LightSIP. Alan's presentation (in Power Point).
- Rolf Rolfsen (Stateoil) gave an update on the status of WIME.
- Jeff Tanner (Schlumberger/OMNES) described the use of standards-based messaging using X.400 and SMTP/MIME as gateways and/or backbone deployment solutions.
- John Broderick (BLM) gave an update on the WRAP-IT initiative. The intent of the project is to improve the efficiency of regulatory systems and data sharing. The ultimate objective is to eliminate paper reporting forms. He asked for oil companies to join the effort. Mike Zeitlin (Texaco) encouraged BHP and other members to proceed with Take-Up. He told the members that establishing the first data store is hard work, but the second one will take a fraction of the time. He suggested that oil companies choose an asset champion and keep the users involved.
In Houston, Ian Shaw (BHP) gave an impromptu short talk about BHP's plan to work with the POSC specifications. BHP wants to see whether they can benefit from Take-Up, i.e. implementing data stores, using data exchange / loaders, and running applications directly.
In Nottingham, there were ten Podium speakers:
- Ole Snerte (Norsk Hydro) gave an update on the Norwegian POSC Forum. Norwegian companies want to purchase POSC-compliant products. They do not want to build such products. Instead of just waiting, it was decided to form a Norwegian POSC Forum. It was started in May 1996 by Statoil, Norsk Hydro, Saga, and NPD. Suppliers will be invited to join at a later time. There are quarterly meetings. The intention is to cooperate on POSC Take-Up projects, to establish common understanding, to highlight areas for cooperation and for proposals, to prepare before POSC member meetings, to achieve a common purchasing strategy, to maintain focus on POSC - POSC/CAESAR integration, and to provide POSC awareness for oil company managers.
- Gerard Huard (Elf) gave an update on the status of LightSIP. Gerard's presentation (in Power Point).
- Nigel Goodwin (Essence) gave an update on the status of WIME. Nigel made this presentation for Rolf Rolfsen (Statoil). Nigel encouraged continued work toward convergence between POSC and POSC/CAESAR. He spoke about the close relationship between drilling activities and both the geoscience and facility domains. He spoke of the need to support large populations of "part" classifications for drilling. He said that he was encouraged by recent work and he encouraged continued cooperation.
- Rick Atkins (Schlumberger/OMNES) described the use of standards-based messaging using X.400 and SMTP/MIME as gateways and/or backbone deployment solutions.
- Gabriele Mormino (Agip) gave a presentation about a migration project (called Production Data Service Application Conversion Project) in which a proprietary data store was replaced with a POSC Data Store using an Agip-defined Epicentre subset. The existing system used object-oriented technology, Smalltalk, and the model-view-controller approach. The only change to be made was in the data store to go from a proprietary data model to POSC. The scope of the work encompassed ten source entities and sixty attributes. This data mapped into 32 Epicentre entities and 43 relational projection tables. Very few changes were required in the mapping layer to support this change.
Questions and Answers:
- Pierre Chevalier (Cap Gemini) asked about performance. Gabriele said that performance was reasonable and that more performance testing was being done. Several people expressed the desire to see performance tuning experiences shared among the POSC members.
Nottingham ACTION ITEM F1:POSC will help disseminate relational projection tuning information and will provide such information in future releases.
- Kitty Hall (Ark Geophysics) gave a presentation a recently approved gravity and magnetics project involving BGS, PECC, and Ark Geophysics. The objectives are to develop gravity and magnetics extensions to the POSC specifications for submission to POSC and to build a gravity and magnetics industry data store. Oil companies will benefit by having improved data quality, by being able to outsource data management, by having improved data access, and by having access to BGS data. The UK DTO Offshore Supply Office is funding 50% of phase 1. Sponsors will be sought for phase 2 which is planned to populate the data store and to develop tools.
- Ross Davidson (Oilfield Systems) gave a presentation summarizing the work of the Petroleum Science and Technology Institute (PSTI) which has recently rejoined POSC as a member organization. PSTI has 36 oil company sponsors. The mission is to obtain access to new technology and ideas. One effort is to build a knowledge base of international E&P projects. See PSTI's Petroleum Technology Research Information System (PTERIS) Web Site at http://www.psti.co.uk/.
Nottingham ACTION ITEM F2:POSC will clarify membership relationship between PSTI members and POSC.
- Barrie Wells (Robertson) gave a presentation about the Regional Exploration Using Epicentre project on behalf of Jim Harris. The project involved the collection of diverse geological data. Needed extensions were made in the "spirit" of Epicentre, e.g. for modeling facies model rock features with geological processes. The emphasis was not on data management. The experience was a good way to "kick-start" POSC Take-Up.
Questions and Answers:
- Tom Gulman (Shell) asked whether the extensions had been submitted to POSC. Barrie confirmed that Paul Maton has a report on the project and that the specific extensions will be provided to POSC.
- Dan Schenck asked how the data was collected. Barrie explained that various DBMS tools were used to populate and access data, e.g. data was fed to ArcInfo for mapping.
- Stewart Robinson (UK DTI) gave a presentation about the UK DTI's involvement with POSC and standardization as it relates to solving a real business problem, i.e. digital government data collection. The first stage of work is to use the semantics of Epicentre. Access will be Web-based. Fourteen basic well header data items have been selected and will be exchanged using an ASCII format. In the future, oil companies will submit data to the DTI using this format. Stewart's presentation (in Power Point).
- Jim Fairley (Codd & Date) gave a presentation about a data model mapping methodology.
Michael reported on the results of a POSC study performed by Marc de Buyl to assess opportunities for POSC Take-Up. The main criteria for the study were motivation, readiness, and funding. (The presentation slides are not posted here. Detailed results are POSC Confidential. Contact POSC for further information.
Nottingham Questions and Answers:
- Stewart Robinson (UK DTI) raised the issue of "data loading". Dan Schenck noted that other than problems with the POSC specifications, data loading tools must come from the commercial community. Stewart asked how POSC allocates resources to the various work efforts. Bill Bartz replied that the strong signals from the members are sensed as in member meetings such as this one. Keith Steele commented that the need for data loading solutions (tools, expertise, etc.) is a market activity for vendors. Hans Stokka observed that the PEF exchanges presume the use of Epicentre at both ends. In contrast, many exchanges involve existing industry formats, such as SEG-Y. Barrie Wells noted that the earlier Industry Pilot Project did a lot of work on data loading.
- Michael Ring clarified that POSC's goal was to be a neutral beacon in the development of national data repositories to help avoid the creation of incompatible solutions. Bill Bartz mentioned activity in Bolivia, Argentina, Brazil, and the Gulf of Mexico. He noted the strong interest of the World Bank to help fund such projects.
No specific topics were proposed in Houston.
Nottingham Suggestions:
- Revisit June Recommendations Not Yet Addressed
[Editor's Note: These include: Process parameter handling in Epicentre, Data Base Services, Technology Day, Integration Solutions, Training Portfolio, How to Manage Data Warehousing]
- Release Management Process Update
- Data Exchange Workshop (value, experience, products, loading, etc.)
- Next Release Progress Update
- Shell: Data Management as a Process
- Interoperability and Business Objects Update
The results of the POSC - POSC/CAESAR joint study called "Hot Chocolate" will be posted on the POSC Web site this month.
In Houston, the level of interest for a Houston-based interoperability / business objects workshop was raised. At least ten attendees indicated interest in a Houston workshop on interoperability in October.
Houston ACTION ITEM E2: POSC will set up a new email discussion list (next-release@posc.org) to support the process of defining the scope and content of the next release.
Nottingham ACTION ITEM F1: POSC will help disseminate relational projection tuning information and will provide such information in future releases.
Nottingham ACTION ITEM F2: POSC will clarify membership relationship between PSTI members and POSC.
POSC Houston and London
10777 Westheimer; Suite 275; Houston, TX 77042; USA; Phone +1 713 784-1880; Fax +1 713 784 9219
Status Four, Status Park, 3 Nobel Drive, Hayes, Middlesex UB3 5EY; UK; Phone +44 181 607 5950; Fax +44 181 759 0465
Email: info@posc.org