from the department of "the end users have their say..."

Mr. Masood Khan, from SABIC, presented a fascinating case study of an extremely fast track project to do an upgrade to a NIGC (that's National Industrial Gas Co. or GAS, from Saudi Arabia). Mr. Khan is the site project superintendent at the NIGC complex in Jubail, KSA. There are five gas plants, producing Nitrogen, Oxygen, and assorted noble gases in the Jubail site. Phases I & II were very old, with conventional instrumentation later converted to DCS. Phase III is a fairly large site, 1200 T/day of Oxygen, producing Nitrogen, Oxygen and Argon. Phase IV and Phase V are newer and larger. The project had several goals: Migrate Phase III from Yokogawa Micro-XL DCS to the latest version of Centum CS3000. Replace PLCs from Siemens and Rockwell that were supplied as part of packaged skid mounted systems. Migration of PLCs suppied with the packaged systems to the DCS. Control of all packaged units from a single DCS platform. Improve system reliability, and reduce maintenance cost Complete control system integration of existing Phase II, III, IV and V DCS systems. "We should," Mr. Khan stated, "be able to control any process in any plant from any operator screen." The project replaced all the Micro-XL controllers, I/O, marshalling cabinets, and some terminations, with new. The project replaced 6 Allen-Bradley PLCs with DCS controllers, and migrated control software to the DCS. The A-B PLCs controlled the Nitrogen conpressors, expander turbines, ESD and molecular sieves, the LOX and LIN Backup Storage Units. The project replaced 3 Siemens PLCs. The Siemens PLCs were running the Oxygen and Air compressors. There were approximately 1200 I/Os, mostly from the PLCs. The project started 10 November 2005, and finished 10 May 2006. Shutdown was predicted to take 10 days, and the total run time of the project was expected to be 24 weeks. In fact, Yokogawa took 17 days to remove the old DCS, PLCs, and replace them all during shutdown. This meant the project took 25 weeks instead of the 24 estimated. Mr. Khan presented some blunt "Lessons Learned." 1. Due to YME's limited field engineering resources, this project was uphill from the start. 2. YME didn't timely appoint dedicated site engineering teams to do the site survey--many drawings were done during shutdown. 3. Gap between ITB issuance and kick-off meeting was too short for YME to actually grasp the project scope. 4. The shutdown schedule produced by YME lacked alignment with the site subcontractors. 5. The gap between the DCS FAT and the start of shutdown was only 3 days"”far too short to establish an effective site execution team. 6. Regular system engineers can't be effective project/construction engineers, so YME mobilized engineers with project experience from elsewhere in the region. 7. Ineffective lines of communication between YME and their sub-contractor led to delays in site execution. Notwithstanding these "Lessons," Mr. Khan noted that every objective was achieved, or is in progress (the integration of Phase II, III, IV and V requires individual plant shutdowns which are still ongoing). Mr. Khan then charitably invited a Yokogawa Middle East representative to discuss whether Yokogawa had in fact learned the lessons he described. Mr. Rizwan M.. Shah, the executive vice president of sales, pointed out the following: -- the project described has led directly to the formation of YSA Services, within Yokogawa Saudi Arabia, with a complete design and construction team, already in the Kingdom so that problems with personnel and visa allocation no longer pose problems. -- subcontractors now meet for communication and alignment and scheduling DURING the FAT, not after it. Mr. Shah joked, "When panic sets in, it is like quicksand"¦" The plant came on line without a hitch, despite the last minute confusion. It wasn't the migration that was the challenge, Mr. Khan concluded, but rather the major challenge turned out to be the site related activities. It is important to note that with all the things that Mr. Khan pointed to as deficiencies in YME performance on this project, IT STILL WAS ONLY 7 DAYS LATE! Not bad. Not bad at all.