As technology capabilities and customer requirements have evolved, so have gas turbine testing and validation methods. Mitsubishi Hitachi Power Systems believes the proof of its long-term validation process is in its turbines’ reliability metrics.
Why would one want to validate a turbine design? The answer is that actual operation of new and more advanced gas turbines can reveal issues that are very difficult to predict on the drawing board. As a result, the validation process must focus on the detection of design deficiencies that can be inadvertently introduced when aiming for improved performance.
Gas turbine validation has improved considerably in the past two decades. The previous approach of testing at a “beta” facility has evolved into a well-thought-out verification process at the original equipment manufacturer’s (OEM’s) facilities, and that has reduced clients’ exposure to prototypical issues and the associated correction process. Based on insurance community feedback, more comprehensive validation has resulted in a noticeable decrease in claims associated with new and prototypical gas turbines.
From Shop Test to Plant Validation
Mitsubishi Hitachi Power Systems (MHPS) started performing shop tests on its gas turbines in the 1980s at Takasago Machinery Works. MHPS’s first step to verify turbine behavior and detect design issues consisted of running the gas turbine coupled to an electric generator while using a water break to dissipate the generated power. This approach allowed detection of typical issues that might occur during the first few hours of operation. These include the tendency to experience compressor surge, vibration, blade tip rubbing, high cycle fatigue, and several potential failure mechanisms that take place during short-term operation.
Although this approach represented an improvement to postponing verification of a prototype until installation at a client site, numerous failure mechanisms will not occur until considerable operation time is accumulated. Performing only a few hours of validation essentially means that one is using the client’s site for long-term validation. But long-term OEM validation can be very expensive. The main roadblock to extending validation at the OEM facility for longer periods of time is the cost of fuel, as it can surpass the cost of the gas turbine in just a few months.
MHPS’s solution to this financial constraint was to construct commercial combined cycle power plants with dispatching contracts to the local utility. This concept ensures operational expenses are covered while the plants satisfy the clients’ generation requirements.
Figure 1 describes MHPS’s approach to the long-term validation cycle. This cycle involves a continuous process wherein the original design is developed in close interaction with the research and development (R&D) team. The new unit is manufactured and operates commercially for long periods at the MHPS combined cycle plant (T-Point), where designs endure the rigors of commercially dispatched units.
1. Mitsubishi Hitachi Power Systems’ design and validation process. Courtesy: MHPS
Detailed inspections are conducted, looking for any signs to indicate an opportunity for improvement. Enhanced designs are developed—again in close cooperation with R&D members’ support—and installed back into the unit to be exposed to the same rigorous verification process. This cycle is repeated for as long as it takes—until designers and researchers are satisfied with the resulting reliability.
MHPS’s long-term validation approach for a prototype is typically conducted for one or two years before units are offered to potential clients. Even after this time frame, the T-Point unit remains in commercial operation to fulfill the plant’s contractual commitments for supplying electricity. A new plant featuring the same equipment validated at T-Point will benefit from further enhancements developed during the two to three years of construction of the entire new plant, since those enhancements are phased into the MHPS commercial product with a focus on continuously improving the gas turbine performance and reliability.
This approach was first implemented in the 50-Hz area of Japan, at a plant named K-Point in Kanazawa. It was later applied in the 60-Hz region at Takasago Machinery Works (T-Point) with a 1 x 1 combined cycle plant, shown in Figure 2.
2. The T-Point validation combined cycle plant. Courtesy: MHPS
Motivation for Building a Commercial Validation Plant
Aeroderivative (aero) engines are designed to be compact and light, and their validation can be set up relatively quickly (Figure 3). In contrast, heavy-duty gas turbines involve the transportation and installation of frames that can be more than 30 feet long and weigh 200 to 300 tons (Figure 4). The logistics involved in shop validation of heavy duty equipment are challenging and expensive, especially if all of the effort and resources are focused on the evaluation of short-term effects for just a few hours.
3. Testing of an aero engine. Courtesy: Pratt & Whitney
4. Large frame adds large challenges. Transporting a large frame turbine, here an M501J, poses challenges because of its size, which can be more than 30 feet long, and weight, which can range from 200 to 300 tons. Courtesy: MHPS & GRDA
The engineering desire to perform long-term validation, plus the elevated logistic costs, provided an incentive for MHPS to build an economically self-sustained plant that allows for long-term verification and extension of the validation process once corrective modifications have been introduced.
T-Point validation involves running the gas turbine and auxiliaries arranged in the same MHPS standard power plant configuration that future plants featuring that type of unit will experience. The plant replicates the operating conditions that a future plant will endure for 25 to 30 years. The vibration modes of a turbine-generator train are heavily dependent on several factors, including: the equipment connected; individual equipment inertia and coupling types; bearing span; and support stiffness, damping, and individual thermal growth.
Using a standard electric generator is particularly important, as its mass and inertia is very different from other types of equipment. The design of the enclosure and its ventilation allows for evaluation of external cooling of the unit under different loads and seasonal changes. By running connected to the grid, the electric equipment associated with power generation—including excitation, automatic voltage regulator, and their protections—is also tested when performing long-term validation.
Short-Term Validation Limitations
Many failure mechanisms are highly dependent on time, whereas others depend on time and temperature. Low cycle fatigue, thermal fatigue, creep, and thermal barrier coating deterioration are just a few that are particularly important to understand for the steady operation of gas turbines and require time to develop and detect.
Short-term validation involves operation of equipment under “new and clean” condition, making it difficult to assess the degradation behavior of the prototypical design. Additionally, wear and tear and its resulting performance deterioration can induce combustion instabilities that may not be present under “new and clean” conditions.
The validation of gas turbines can include stringent tests to simulate severe operating conditions that might never be experienced in the field. These tests should demonstrate that the new turbine can withstand the strain implied in those tests. However, potential clients and insurance companies are usually concerned with the permanent effects such tests can have on the hardware and are hesitant to purchase these turbines. The gas turbines validated at T-Point belong to T-Point and are never sold to a client as a new unit.
Data vs. Demand Conditions
Data collection and analysis has been an important keystone for the advancement of science and engineering. The enormous progress in computer power, simulation techniques, and sensor technology experienced in the last decades has further improved our industry’s understanding of individual failure and deterioration mechanisms.
Large amounts of data can be collected during testing or operation of new gas turbines, enhancing the designers’ understanding of different potential failure mechanisms; however, the complex stress-loading of components, especially the hot gas path parts, still makes it difficult to fully correlate data collected and the resulting distress and potential failure mechanisms.
Unfortunately, this limitation cannot be simply overcome by increasing the volume of data collected and analyzed. This fact becomes evident by the complexity of root cause analysis (RCA) of gas turbine failures. RCAs typically arrive at a list of possible scenarios rather than a clear cause, despite the availability of damaged parts, ample data collected before and during the incident, and the intervention of talented and experienced metallurgists and fracture mechanics specialists.
There is no doubt that gas turbine design has benefited enormously from recent computational enhancements, but the verification of the components’ design resilience to complex failure mechanisms, and their durability under prolonged operation, can only be determined by long-term exposure to sustained operation under different demand conditions.
The Next Validation Plant Evolution
At the time K-Point and T-Point went commercial, the computational power mentioned above was not as advanced as it is today. Application of more-powerful computational systems in R&D facilities and continuous efforts to retrofit effective data-gathering systems to the existing T-Point validation plant have considerably improved MHPS’s understanding of complex stress interaction and potential failure mechanisms. However, for MHPS, the conclusion drawn more than 20 years ago remains unchanged: The final verdict regarding parts resilience to time-dependent failure mechanisms and durability requires long-term operation, regardless of the volume of data collected in labs or during short-term controlled operation.
MHPS G-Series turbines were the first to be subjected to the comprehensive validation process at T-Point. Today, the G fleet has achieved a five-year rolling average reliability of 99.1%, according to the well-known third-party company Strategic Power Systems (SPS). The recently added J-Series gas turbine fleet, which now has 17 units in commercial operation around the globe, also went through the rigorous validation process. To date, the M501J fleet has accumulated over 220,000 actual operating hours and, more importantly, has already achieved a reliability record matching MHPS’s M501G fleet of over 99%.
Building on this long-term validation success, a new state-of-the-art combined cycle plant is currently under construction adjacent to T-Point. This new combined cycle plant will incorporate state-of-the-art computational tools, but above all, it was designed based on MHPS standard plant configuration and will also be connected to the grid for extended periods of verification. This large investment constitutes a clear demonstration of MHPS’s conviction that both short- and long-term validations are required to reduce risk of failures and enhance reliability. ■
—Carlos Koeneke, PhD is VP of Project Engineering at Mitsubishi Hitachi Power Systems Americas Inc.