5SEG Technical Standards Committee Meeting (SEG Denver Oct 2004) SEG-D Review SubcommitteeAlan Faichney had proposed that this be disbanded due to lack of volunteer effort.Committees were set up two years ago to review and they reported. Then working committees were then set up, however no progress was made.General feeling that this should not be left to drift and it was proposed that we should go back and solicit input from manufacturers and users and any other interested groups.Previous revisions to SEG-D were driven by need as a result of manufacturer changes to equipment. (e.g to 3590 tapes)Oil companies would probably support changes that would have value but not just for a few header changes.The new IBM format tapes might be a big enough change to drive this?If a new format was to be defined it was important that all stages of the processing business right through to interpretation must be represented. Ted Mariner said it was his responsibility to make sure the support was found and given!Jill Lewis said this needed a small core group of people working on the format with a wider review group. This was the model used for SEGY Rev 1 and had worked well.
6SEG-D vs Tape Standards SEG-D rev 0 (1975):The nine track one-half inch tape is retained, but extension to 6250 BPI is permitted as long as it is compatible with the ANSI standard (ANSI 1973a, b, 1976).SEG-D rev 1 (1994):Providing provisions for using 1/2" square tape cartridges. ( ANSI X ).SEG-D rev 2 (1996):Since Rev 2 is intended to handle higher density tapes, acceptable media is expanded to include: 3490/3490E, 3590, D2, and D3.SEG-D rev 3 (EAGE Madrid June 2005)It was decided that within the current timeframe and the general desire to take advantage of ultra high capacity media a revision of SEG-D to rev 3.0 is appropriate at the present time.
7Problem description The stipulation in the 1975 standard that “no more than one line of seismic data ispermitted on any one reel” becameimpractical long ago.SEG-Y rev 1 (May 2002)SEG-D rev 3 (Nov 2005?)
8Example (marine) - 10 x 6 km streamers (12.5 m group length) - 40 km line length (=1600 shots per line)- 10 sec shot interval- 8 sec record length, 2 msecFormat: SEG-DAmount of data per shot: ~ 50 MBAmount of data per line: ~ 80 GB3590E: USD 20? / GB / : $/ GB3592 vs 2 (TS1120): USD 125? / 500 GB : $/ GB
10EAGE Madrid June 2005 Question SEGD Rev 3.0 or SEGE – vote It was decided that within the current timeframe and the general desire to take advantage of high capacity media a revision of SEGD to rev 3.0 is appropriate at the present time.Concerns driving the requirement for a Rev 3.0The receiver count is ever increasing and when this is multiplied by 4 (for four component recording) then the present limit on the number of channel that can be recorded becomes an issue.+++
11EAGE Madrid June 2005Critical Items and volunteer companies who will present the working group (reflector site members) with their recommendations:4C Western Geco / Veritas / Sercel / IOChannel Set DescriptorsXYZ+Time - Concept / Saudi Aramco / Sercel /SPS CompatibilityFile Numbers and Multiple Lines - Troika/ Verify / Western GecoBlocking + Non-blocking & Separate LayersThe challenge to the group is to come up with a draft set of documents to be handed to Ted Mariner by 1st October.
12SEG-D revThe subcommittee consisted of the following individuals:• George Wood Western Geophysical / Chairman• Phil Behn Input/Output• Claes Borresen PGS• William Guyton Western Geophysical• Louis Miles Syntron• Dennis O‘Neill Geco-Prakla• Sut Oishi Shell• Tony Scales Sercel
13Rev 3 proposed ’changes’ (1:39) Issue Status Resp. Deadline CommentUpdate ftp server Bob Nov 15’thNew input All Dec 15'th ….Chapter 1 (Intro) Jill Jan 1'st ….Chapter Jan 1’st1: ….2: …..….
14SEG-D rev 3 App F: Maximum Block Sizes The table below indicates the maximum allowable block size for accepted types of media. It is expected that this table will need to be updated approximately once per year.Device Type Maximum Block SizeKilobytes3490, 3490E KilobytesKilobytesIBM (Jaguar-1) 2,097,152 Bytes (2 Megabytes)IBM TS1120 (Jaguar-2) 2,097,152 Bytes (2 Megabytes)DST1, ,840 BytesRedwood KilobytesSTK 9940B 256 Kilobytes1/2” Round Tape KilobytesKilobyte is defined as 1024 bytesMegabyte is defined as 1024x1024 bytes (=1,048,576 bytes)
15SEG-D rev 3 App C: API codes Rev 3 (One simple question: Address correct?)POSC 24 Greenway Plaza Suite 1000-B Houston, TX 77046 USA telephone faxRev 2 (No longer correct)American Petroleum InstituteExploration and Production Department1220 L Street, N. W.Washington, D.CPhone: (202) 682‑8000FAX: (202) 682‑8426
16Appendix A: Manufacturers of Seismic Field Recorders 40 Geo-XSuite 900, 425 1st St SWCalgary, Alberta, Canada T2P3L841 PGS ???42 ???43 Hydroscience ???
17Appendix G: Multiple SEG-D header blocks Proposal (If SEG-D header is longer than maximum physical block size).N-1 header blocks with fixed size, block number N without padding:By reading and decoding the very first 64 bytes (GH #1 and GH #¤ 2), it’s possible to calculate the total SEG-D header length. If block size and total SEG-D header length are known, it’s possible to calculate number of physical tape header blocks.Feedback wanted.
18Rev 2 BOT Tape label 128 bytes / Required General Header # 1 / ” General Header # N / Optional (N>2)
20SEG-D rev 3 proposalsWe recommend to introduce 5 new required General Header blocks (#3,#4,#5,#6,#7) due to multiple lines.Vessel/crew identification (with abbreviation?) ASCII*32 ?Area name (with abbreviation?) ASCII*32 ?Client name (with abbreviation?) ASCII*32 ?Job name (with abbreviation?) ASCII*32 ?Line name ASCII*16 ?Is it possible to get an agreement how to generate line name?[Area:2][Client:2][year:2][line type:1][linenumber:4]-[sequence:4]WGecoTape Label Serial number: [vessel:2][job:4][reel number:6]
21Proposal for discussion: New required GH blocks #3-#7 1 byte block ID, 0/2/4 bytes abbreviation (ASCII) bytes ASCII and 1 byte statusID Abbreviation Comment ExampleSE Crew Id Sea ExplorerNS Area Name North Sea - StatfjordST Client StatoilJob WesternGeco 8104Line Name STStatus Codes:0x01 OK0x02 Not OK, e.g. data in this header block are not valid0x03 Irrelevant