The increasing pressure to reduce emissions, while at the same time maintaining good economics and high reliability is placing onerous demands on gas turbine technology. The indications are that the challenge can be met but it is proving difficult. There have been traumas along the way and there are likely to be more. Glossing over the difficulties is not helpful. There is no shame in admitting problems, indeed problem solving is what engineering is all about, particularly when you are pushing at the boundaries of a technology and stretching the ingenuity of turbine designers to the limit.
Most gruelling is being at the head of the pack, pioneering new technology which promises much but which is relatively unproven. When it comes to aeroderivative gas turbines using dry low emissions (DLE) technology the Belgians are very familiar with the ups and downs of being in this “fleet leader” role. At Modern Power Systems’ recent conference on power plant operations and maintenance (held in Brussels, Belgium,12-13 April) a paper presented by Jan Vanoudendycke of Electrabel (and co-written with Jean-Luc Vandesteene of Tractebel Energy Engineering and Marc De Witte from Electrabel) gave a refreshingly and commendably frank account of the Belgian experience.
The choice of DLE arose from the “expected cost increase of demineralized water” and from “a general request from the Belgian authorities that the best available technologies should be used to limit detrimental emissions.” By early 1999, eight GE LM6000 DLE machines and one LM2500 DLE were in commercial operation in Belgium, principally in cogeneration applications.
The Gent district heating cogeneration plant of public utility SPE was the first commercial installation in the world of an LM6000 with the DLE combustion system (designated the LM6000PB). The machine was installed as part of a natural gas fired combined cycle plant to replace old coal fired boilers and a condensation steam turbine. The power plant was located in the centre of a town; it had to be small, highly efficient, require little cooling water, and generate minimal emissions. Despite the anticipated higher maintenance costs of the aeroderivative technology compared with an industrial gas turbine, the LM6000 option came top in the economic evaluation and SPE “decided to take the risk of a world premiere.” The Gent plant was commissioned in 1995. Low emission high efficiency competitors such as Rolls Royce’s Trent and ABB’s GTX 100 have emerged since but were not around at the time.
The DLE combustion system at Gent was found to be “complicated and sensitive”, but generally performance indications looked promising. There were initial problems with combustor heat shield cracking, but these were expected, having been encountered on a test engine before the Gent unit was installed. The cracking led to an early engine changeout. Subsequently new premixers, combustor and HPT2 blading was fitted and eventually the combustor was replaced with one of a new design, with additional cooling holes. But this is the sort of thing you expect when you are working with relatively unproven technology.
Further LM engines were ordered: one LM2500 package for Izegem (commissioned July 1996) and three LM6000 packages for Zeebrugge, Lanaken and Geel, commissioned December 1996, January 1997 and November 1997 respectively. These were all cogeneration projects. Geel was another world premier: the first LM6000 to be equipped with the improved DLE combustion system, designated LM6000PD (all subsequent machines have been of this type). In addition one spare engine was purchased and three more LM6000 packages were ordered, for repowering schemes, two for Langerlo (commissioned in March 1998) and one for Ruian (commissioned June 1998). Subsequently another LM 6000 package was procured for the Aaalst cogeneration plant (commissioned December 1998). In the first half of 1997 “serious problems” were encountered in the Zeebrugge and Lanaken engines, including LPT5 blade (last stage power turbine) failures and fast combustor deterioration. Delays in decision making and late or wrong spare parts deliveries compounded the problems and led to very extended outages. To make matters worse, in May 1997 abnormal vibrations were detected on the Gent engine. The exact cause was difficult to trace, so it was eventually decided to disassemble the machine. This revealed that the origin of the problem was loose broken parts from a failing air-oil seal baffle on bearing No1 in the inner casing of the LPC rotor.
At times, in the first half of 1997, there was misery and even panic said Jan Vanoudendycke, “but we survived and got through it.” Unfortunately the first few months of 1999 have also proved trying. For example, in January 1999 it was found that the bearing of an inlet gearbox at Langerlo had failed. During installation of a spare engine a problem with final connection of the oil system led to core engine damage, requiring installation of a lease engine. In March 1999 at Aalst a similar bearing problem to that encountered at Langerlo was discovered. A lease engine has been installed and root cause analysis is underway.
A range of other problems are mentioned in the conference paper and, understandably, the Belgians are planning to put more effort into condition monitoring of their plants in the future. But Jan Vanoudendycke remains a firm believer in the technology, pointing out that, for the matured systems, NOx emissions (“well below the guaranteed 25 ppmvd, close to single digit”) and power output (over 49 MWe at -5ยบ C intake air temperature) are better than expected.
A number of additional LM6000PD packages are planned in Belgium, including six in Antwerp harbour and most recently two for the Solvay Jemeppe chemical plant. It will be very interesting to see how the performance of the Belgian fleet develops and how we are progressing up the learning curve of low emissions turbine technology.
