Proactive management of GT parts life key to controlling maintenance cost

By Salvatore A Della Villa, Strategic Power Systems Inc

The gas-turbine-based segment of the electric power industry is evolving rapidly , considering most change in the electric sector is measured over a period of decades. Only a few years ago, gas was under $3/million Btu and large combined-cycles were being designed for base-load service. F-class GTs dominated combined-cycle orders and longterm service agreements (LTSAs) championed by the OEMs (original equipment manufacturers) were the risk-mitigation strategy of choice by owner/operators.

The plan looked perfect.

But then the wheels came off. In short order, California’s new competitive electricity market— thought by many to be the model other states would follow— collapsed, gas doubled in price, and environmental rules became more challenging.

Today, gas prices are at historic highs, emissions of NOx and CO are approaching zero, and units designed for base-load service are operating in a cycling mode. This cyclic operation causes GT parts to wear out much faster than the owners originally planned. Profit margins are razor thin and LTSAs for plants with uncertain operating requirements may no longer be an effective risk-mitigation strategy. New risk-management approaches are needed to survive the difficult environment.

It is important to note that approximately 70% of the total maintenance cost through the full life cycle of any GT is spent on new parts to replace those that have aged or been damaged to the point that they cannot be refurbished and returned to service.

What this means is that the winning bidder for an electricity supply contract—assuming the competitors are two (or more) similar units with comparable gas contracts—will be the facility that best manages its physical assets. This economic reality is strong motivation for understanding what is required to get the most, in terms of lifetime, out of hot-gas-path (HGP) and other critical parts relative to their empirical life limits.

Parts life tracking and criticalparts life management must be performed with diligence, accuracy, and with an understanding of the economic consequences, to achieve realistic and acceptable total life-cycle costs.

The new paradigm. The issues that constrain the ability of owner/ operators to implement effective repair and refurbishment strategies for high-cost parts has been exacerbated, as never before, by the operating profiles of their units (starts, stops, hours), temperature and load excursions, trips at various loads, fuel (natural gas vs distillate), water (NOx abatement and/or power augmentation), and the inherent design and metallurgical characteristics of “high technology” parts (advanced materials, cooling, coatings).

The influence that these process variables have on the expected life of HGP parts has resulted in the development of “factored” algorithms by OEMs to more accurately reflect ageing. Maintenance schedules are based on so-called factored starts and factored hours rather than actual starts and hours.

The sidebar and tables explain the significant impact that duty cycle has on maintenance intervals. A look at the financial ramifications of the new operating paradigm for an F-class GT is instructive. Critical information for decision-making: A new set of first-stage rotating blades costs about $3-million; fall-out rates (blades scrapped) range from about 25% to 30% for units operating in a cyclic mode; refurbishment cost for a complete row of worn but sound first-stage blades is in the neighborhood of $500,000. Clearly, there’s an economic incentive to recognize and act on the importance and value of parts-life tracking.

Data collection is the first step in creating information that enables accurate assessments and decision-making. Operating plants are inundated with data collection and reporting requirements that are both burdensome and costly. The perceived cost/benefit of recording detailed information often is discounted, if not under-valued. This perception must change for effective and meaningful parts-life tracking to be performed efficiently, accurately, and cost-effectively.

GT operating profiles changed markedly from 1995-1999 to 2000-2004
A review of information available from the ORAP® system (background in companion sidebar) shows that the operating profiles for key classes of gas turbines (E, F, and aeroderivative) have changed significantly between the periods 1995- 1999 and 2000-2004.

Table 1 reviews key empirical ORAP information—including service factor, service hours per start ratio, and average number of starts annually for the two five-year periods.

Data for 2005 through mid July, which covers a significant portion of the summer peak, show that trends for service factors and service hours per start noted for the two periods studied continue in all but one case (an uptick in service hours per start for E-class machines; however, starting frequency for this type of GT continues its downward trend).

Other observations:

  • There is a general downward trend in the number of annual service hours for both E-class and aero machines. For the E, average service hours declined by 22% (approximately 1000 hours) from 1995-1999 to 2000-2004; for aeros it was 33% (about 2000 hours). There might have been a decline for F machines as well had the reliability of those units met expectations during their early service years (1995-1999).
  • Units across all classes are operating fewer hours per start: Es down 13%; Fs down 23%; aeros down 46%.
  • F-class and aero engines averaged about 20 more fired starts per year in 2000-2004 than they did in 1995-1999. By contrast, starts declined for the older E-class units. The foregoing suggests a change in GT duty cycle or operating mission. A cyclic profile is in evidence. Because the unit’s operating profile is one of the key variables influencing maintenance intervals and parts life, it is important to assess whether the changes in operating trends have impacted historic maintenance intervals.

Table 2 reviews empirical ORAP information for the interval between hot-gas-path (HGP) inspections on both an hours and starts basis, and the time it took to perform the maintenance (elapsed or unavailable time). Note that 2005 data were excluded from this table because it represents a partial year and planned maintenance had not yet been fully performed or reported by press time.

Important observations:

  • Intervals between HGP inspections for E- and F-class GTs have increased on both an hours and a starts basis.
  • There has been a substantial improvement for F-class turbines in the inspection interval—more than 50% based on hours, about 90% based on starts.
  • Characterizing the 2000-2004 values on an “equivalent” basis to demonstrate the potential impact of the OEM’s “maintenance factor” on the HGP inspection interval indicates that today’s cyclic operating regime clearly has a significant influence. Specifically, the so-called factored hours used for determining the maintenance interval are double (2.1 to 2.2 times) the number of actual operating hours; factored starts are five to six times the number of actual starts. Note that the table presents the actual hours and starts.

Maintaining comprehensive parts tracking information is a basic “good management” practice in any situation where the owner/operator bears any of the risk relative to the cost of replacement parts. However, even in cases where the risk has been mitigated through an LTSA contract the owner/operator should either have an independent information source or work with the OEM to understand and maintain the data collection and analysis results. Over the long term this information resource will be required to justify any decision regarding parts refurbishment or replacement.

The maintenance of this information is not just a concern for ongoing operations, but is also a business requirement. In the event that the asset is put up for sale, a buyer will typically require a comprehensive accounting of the expended life on all critical serialized parts to enable an accurate assessment of value. The bottom line: It is important for you to understand how ageing factors are impacting plant costs and to know the life remaining in critical parts. This understanding can only be achieved through a comprehensive and disciplined approach to collecting and storing all relevant parts-tracking information.

Access to the process control data directly from either the unit level control or the DCS provides opportunity for productivity benefit and payback based on the ability to accurately compute and time-stamp the aging factors based directly on the unit’s operation. The process data permit the development and retention of ageing characteristics.

Associating computed ageing- factor data with the inspection data, on/off records at a serial-number level, and repair/ refurbishment data (including as-found and as-left condition) is the basis for maintaining the current configuration or pedigree of the unit at a parts level. For a fleet, it provides the ability to manage the pool of available parts or shared inventory in a cost-effective manner—directly associating the parts history and usage profiles across the various units.

What is ORAP®?

ORAP (for Operational Reliability Analysis Program) is an automated system for monitoring and reporting the RAM-D (reliability, availability, maintainability, durability) of power and process plants equipped with gas and/or steam turbines. Data currently are tracked on more than 2000 units worldwide.

Support for this activity comes primarily from turbine manufacturers. The relatively small number of plants not covered by manufacturer funding, or those that do not want to share operating data with their turbine manufacturer, can participate for a modest fee. Details on participation are available at www.spsinc. com.

Information compiled in the ORAP system covers various applications, duty cycles, and plant arrangements for both simple- and combined-cycle facilities. Data are collected and reported from the “bottom-up,” enabling engineers to see the impact of a component failure on the system it serves and on the plant as a whole.

EPRI’s (Electric Power Research Institute, Palo Alto, Calif) standard equipment codes, developed by SPS, are the basis for reporting system and component outages. These codes provide reporting uniformity across equipment types and they are fully compatible with the European KKS system codes.

There are several methods for tracking parts. They differ in level of sophistication and implementation cost, ranging from simple spread-sheet applications to sophisticated enterprise resource planning (ERP) tools. Time, complexity, and cost are key variables. Your objective should be to select an approach that maximizes automation and minimizes manual labor.

The method preferred by Strategic Power Systems is to automatically extract data from the unit control or DCS and deliver it to the company’s ORAP database engine (see sidebar). ORAP transforms the real-time process control data into ageing factors using OEM- or owner/ operator-supplied algorithms, associates these ageing characteristics directly with parts at a serial-number level, and provides the owner/operator access to the parts history over the Internet through an encrypted secure connection.

The user has the opportunity to increase the value of the database by adding information on parts condition, refurbishment notes, before/after photos, etc. This is one way of capturing valuable institutional memory. The tools for implementing this enhancement are relatively easy to use.

Don’t forget the aeroderivative GTs. The focus here has been primarily on large frame machines, but parts tracking and life management of critical parts has become increasingly important for aero engines as well. Changes in duty cycle for these machines (Table 1) have mandated consideration of the cyclic life limits for critical parts. Tracking the cyclic life has become more important as manufacturers begin to place absolute limits on the life of critical, capital engine parts for equipment operating in land based applications. CCJ OH

Salvatore A Della Villa Jr, president and CEO, Strategic Power Systems, Charlotte, NC, has 30 years of experience in information technology and reliability engineering, specifically focusing on power generation and industrial processes. Under his leadership, ORAP has become the largest global database of its type serving power and industrial plants. SPS emphasizes the importance of monitoring the operational performance of critical equipment. It offers a family products and engineering services focused on the capture and processing of plant data into information of strategic and tactical value.