Tulsa — The rate of material degradation of geothermal steam turbines makes modernization a wise choice for most existing facilities. The extreme conditions of excessive moisture together with sulfides and chlorides in the steam means that wear and tear take place at a far faster pace than in traditional steam turbine environments. Although great strides have been made in reinjecting geothermal fluids back into the formation, steam-field degradation can set in soon after startup and often prior to the five-year mark.
The first indications come in the form of performance declines, which are typically in the 5 to 20 percent range. The chief cause is a decline in resource conditions resulting in a commensurate decline in turbine inlet parameters. The damage to system components by geothermal fluids and moisture, as well as the influence of contaminants, exacerbates the problem. Therefore, whether you are a considering developing a new geothermal power plant or are an existing owner preparing for your next turbine overhaul, making the correct decisions on turbine attributes can provide significant improvements in reliability, efficiency and total cost of ownership.
During one project completed a decade ago, for instance, the original output of 23,800 kW was increased by more than 10 percent to 26,325 kW. At the same time, upgraded materials and design elements applied during this re-rate enabled the turbine to eliminate recurring wet steam erosion, erosion/corrosion and corrosion-related component cracking problems, thereby improving unit reliability and availability. The new steam-path was optimized to the then-current steam-field conditions, which had changed over the unit’s 10-year operating life. Ten years later, as the unit enters its third decade of operation, this same owner is contemplating a second re-rate, this time adding a new well for more flow and planning to boost output yet again.
In another example, an owner redirected steam from different wells with different pressures to different units. By doing this, one unit’s output was improved by directing high pressure well flows and another unit’s output was increased by directing lower pressure well flows. In the former, the high pressure (HP) section was reengineered and the unit was optimized for a higher inlet pressure. While the original rotor and blade rings were retained, new design HP rotating blades were installed as well as an improved design for the HP stationary blades. The low pressure (LP) inlet was blanked off, too, to route LP steam to another turbine. This alone boosted output from 30 MW to 35 MW. For the unit receiving the lower pressure steam, the steam path was optimized for the lower pressure and higher flow. For that unit, output increased from 30 MW to 34 MW.
Such breakthroughs have enabled older geothermal turbines to achieve turbine efficiencies approaching those of more recent models. By incorporating many lessons learned from a cross-section of operating units and providing geothermal specific, cost-effective designs that lower cost of ownership, solutions can be implemented that give these units a longer life, sustained generation and that offer their owners a profitable future.
As almost all geothermal plants have been through multiple outage cycles, many aspects have been identified for improvement. A number of techniques and maintenance best practices have been imported from the fossil low pressure steam turbine arena. Stress reducing features, for instance, have been extensively derived from this field. The use of titanium for its high strength, low mass and excellent corrosion resistance properties has been borrowed from the nuclear industry. Materials from mature gas turbines such as Inconel are also gaining acceptance.
With the exception of a few direct steam applications such as The Geysers in California and Larderello in Tuscany, Italy, geothermal steam is processed and enters the turbine in a dry-saturated state. Therefore, unlike clean fossil steam applications, even the first stage can be exposed to the damaging effects of moisture. As a result, wet steam erosion and corrosion (don’t forget the sulfides and chlorides in the geothermal steam) can be extreme. These phenomena are commonplace and lead to damage and shortened life in rotor blade wheels, rotating and stationary blades, casing diaphragm fits and horizontal joints. In short, anything in the steam path can suffer heavy damage from erosion and corrosion.
Other areas to investigate are the existing steam-path moisture removal features and the turbine drain system. Oftentimes, this system can be improved upon. A key action in older units is to add more moisture removal features. Installing large drains in every stage, for example, can bring about significant improvements in moisture removal and the damaging effects of this moisture on downstream stages. This should be supplemented with drain line orifices to control drain flow, bypasses and valves, as well as instruments to monitor pressure, flow and temperature. With field piping, for instance, the recommended approach is to use external orifice plates to meter the flow during normal operation. Further, isolation and bypass valves typically need to be installed to allow cleaning or replacement of orifice plates while the unit is on-line. Properly sizing the valves and orifice plates is of vital importance. Pressure and flow instrumentation are important for monitoring drain flow rates and moisture removal.
Similarly, consideration should be given to the introduction of replacement diaphragms with moisture removal features in every stage to permit moisture removal further upstream in the steam-path than originally configured by the original equipment manufacturer.
When it comes to weld repair of steam-path components, the geothermal environment demands that welding be done using stainless alloys such as 12 percent chromium (Cr) steels and noble alloys such as Inconel.
The sheer extent of erosion becomes clear to anyone working on geothermal turbine repairs. Rarely is that degree of deterioration seen in a fossil setting. Geothermal steam properties exact a severe toll on turbine reliability and this becomes readily apparent in components manufactured from low-alloy steel or carbon steel. And when casing or diaphragm seal faces degrade, pitting and washing increases exponentially. A wise move during a major maintenance overhaul, therefore, is to add robust and high quality stainless steel inserts in critical casing steam seal areas.
When corrosion pitting occurs in highly stressed rotating blades, the material endurance strength is degraded and can lead to corrosion-assisted fatigue failures. When replacing blades, improving the damping systems with modern features and upgraded materials can significantly improve reliability and extend operating life. (Left, New rotors installed at a U.S. geothermal plant included moisture removal features. Credit: TurboCare)
Learning from Repairs
In numerous geothermal overhauls, TurboCare engineers have used advanced techniques to shore up aging machines against heavy degradation. This has included restoring gland steam sealing areas using submerged arc welding with Inconel and extensive redesign of components suffering from corrosion-related cracking, as well as 12 percent chromium weld repairs of more typical stress corrosion cracking (SCC) on the turbine wheels. Such repairs entail a substantial amount of engineering, often including complete blade and wheel redesign.
Multiple projects in the geothermal industry have revealed a common problem in geothermal turbines, which is inadequate blade damping. Poor damping can lead to corrosion-assisted fatigue failures. This phenomenon can affect both blades and rotor wheel attachments; hence the need for a comprehensive failure analysis followed by blade and wheel redesign to incorporate modern damping features.
Another important realization is that free-standing blades tend to have inadequate damping for geothermal applications. For this reason, blade upgrades are often coupled with 12 Cr wheel weld repair of wheels and machining to fit the new configuration. In one example, the original part had failed within five years. The new version, on the other hand, has run successfully for 10 years based on improved design principles.
In tall low-pressure geothermal blades, best practices in geothermal repairs has been extended to the application of continuous coupling of blades to suppress blade response and lower alternating stress. Modern designs should be applied whenever blades are being replaced; a replacement-in-kind policy is an inefficient economic decision for the owner.
Alternating stress does not only occur in blades. For instance, an inspection at one project revealed a crack in the shaft-wheel radius which had propagated to within 1.5 cm of the bore. A failure analysis consisted of metallurgical analysis, finite element analysis (FEA) and a site alignment survey. Conclusions drawn from these analyses were that failure occurred due to misalignment causing alternating stress and a reduction in material endurance strength caused by corrosion attack. The solution in this case was to sever the shaft end and re-join it via welding, as well as 12 Cr cladding of the radius and complete realignment of the turbine-generator.
In other cases, however, partial repair of the rotor has not been possible. At one site, a new replacement rotor for a 40 MW geothermal unit was required due to the sheer extent of damage and the amount of performance degradation that had taken place. This work also involved an upgrade to the rotor material to combat SCC and a redesign of the first, third and last stage blades for overall improved reliability and heightened efficiency. Such projects, though, require time. Good planning and timely inspections are needed as it can take a year to deliver and install a complete new rotor and engineer the required enhancements to the steam-path to mitigate ongoing moisture and corrosion-related damage.
The industry continues to improve geothermal rotor materials and blade materials and the trend is towards cleaner alloys and, where necessary, increasing alloying content. The economic benefits of materials improvement derived from better reliability and availability far outweigh the incremental cost increase of better materials. (Below, right: Another view of rotors installed at a U.S. geothermal plant. Credit: TurboCare)
Recommendations for the Future
These findings and resulting best practices are based on a long history of geothermal unit overhauls and upgrades that have been performed over several decades. The conclusions are that modern airfoil technology and materials should be introduced into this field to combat erosion and to upgrade aging units and units struggling to be economic. Turbine steam-paths should be optimized for the changes in temperature, pressure and flow experienced in the steam-field resource. After all, the gradual decline in steam-field conditions is not unexpected, but not to address this decline with modifications to the steam turbine will only result in suboptimal investment returns to the owner/operator. Further, as more wells are drilled, there may well be an additional flow to deal with but at lower pressure.
Therefore, in addition to modernizing geothermal turbines to combat the erosion/corrosion environment, it is essential to plan for pressure and flow variation from the steam-field over time. Whether one is sourcing a new geothermal turbine or upgrading an existing steam-path, the design should be optimized to the pressure and flow conditions that currently exist.
When reliability optimization efforts are combined with thermodynamic improvements to address changing resource conditions, the original output can often be recovered and, in some cases, improvements have been attained over the original output.
In other words, your aging geothermal steam turbine can undergo a life extension that up-rates the machine based on the latest technology and practices.
James Beverly is TurboCare’s manager of Geothermal Steam Turbine Applications and has two decades of experience in the industry. He holds a degree from the University of Connecticut.