ARMA services (asset risk mitigation analysis)
Asset Managers working for utilities today are faced with many challenges. Some are a consequence of market deregulation, while others are driven by aging infrastructure. Below is a list of typical concerns faced by the industry as a whole. These concerns generally impact customers asset management strategy.
- Inherent higher probabilities of failure
- Parts obsolescence, restricted availability, and logistics
- Shortage of equipment expertise.
Merger, acquisition, and divestiture activity:
- Increased diversity of fleet age, manufacturer, location, standards etc.
- Merging different maintenance strategies and practices
- Separation of regulated and deregulated market assets.
Load growth and network complexity:
- Increased Independent Power Producers
- Merging different maintenance strategies and practices
- Integrated planning with regional energy market administrators.
Homeland security issues:
- Planning around or removing infrastructure vulnerabilities from potential terrorist activity.
All of the above must be addressed efficiently by optimizing the use of capital to meet shareholder expectations regarding return on assets (ROA) and/or return on investment (ROI). Compounding the problem is a lack of contextual data to measure asset strategy success. In interviews with many investor owned and municipal utilities, operational and non operational data abounds but lacks integration into day to day operations to provide meaningful and actionable information. Some companies have setup teams to pull data from Energy Management, Distribution Management, and Geographical Information systems and combine the data from Computerized Maintenance Management (CMM) and Work Management (WM) systems to allow stochastic analysis of data. The goal being to correlate asset conditions metrics with operational data to gain insight as to when and what type of maintenance should be performed. Some utilities have applied Six Sigma rigor to understand the relationships, only to find that the data extracted from the various systems is incomplete, lacks coherency, or lacks consistency to be useful.
When companies apply significant resources and eventually build a knowledge based system, which typically can take one to two years, the resulting systems are condition based. Yet asset condition is only one part of the total solution to achieve prioritization of maintenance and capital resources. Reliability impact as well as asset condition are required to achieve true optimization of resources. Hence as shown in the figure above, data must come from maintenance as well as planning to develop a comprehensive asset management strategy.
Many believe that since the original system was designed using N-1 or N-2 reliability criteria the importance of linking condition assessment decision systems with overall system reliability is marginalized. The fact remains however, that over the last five to seven years most systems have seen inherent change associated with loading, expansion, modifications in operating practice, etc. This condition can be found in particular in cases where the assets were recently purchased through an acquisition or merger to provide synergies with respect to generation or transmission. Additionally, with the creation of regional transmission and/or energy marketing organizations, reliability planning and related criteria have become a more shared responsibility. There has been a shift in primary focus to transmission congestion relief to create a more efficient energy market. Hence the degree to which one can assume original criteria being maintained is more of a subjective rather than objective evaluation of the current situation. The relative cost of generation must also be taken into consideration to effectively establish prioritizing criteria. A generator step up (GSU) transformer connected to an energy source having one third of the relative cost of energy production compared to another GSU should have a higher priority from an asset risk management perspective. Moreover, a GSU connected to a critical power station strategically located to relieve transmission congestion would also have a higher priority when developing risk mitigation strategies. The question therefore is how does one go about pragmatically sorting through the massive amounts of data available for analysis to achieve a comprehensive asset management strategy?
For transformers, the answer lies in looking at the various asset groups and establishing priorities based on condition and reliability impact of each transformer. Let's focus our discussion to transformers to illustrate the point. No one would argue that transformer assets represent one of the most critical assets in the T&D infrastructure. From the generation and transmission level, transformers play a critical role allowing power to be delivered efficiently and reliably. One could also argue that transformers also represent one of the most expensive asset types found in utility grids today with replacement lead-times that can be between 12 to 15 months.
Transformer asset risk profile
If one looks at the historical installation rates of GSU, transmission, and large distribution transformers, the total population is approaching a critical stage of maturity.
The transformers installed in the early 70s are reaching a critical age where their probability of failure will be increasing exponentially.
Based on these statistics, failure rates are on the increase. The number of explosions has increased and failure rates are expected to increase exponentially as the installed fleet ages. Transformers manufactured in the 1960s to the early 1980s have a failure rate that is 3.75 times the normal failure rate as shown in Figure 4 below.
In review, the current transformer installed based in North America it is conservatively estimated:
- There are 29,000 transformers at transmission levels.
- Average age is 35 to 40 years - increasing at .7 years.
- 9,000 considered critical (based on load, network position)
- Only 1,300 units have on-line monitoring (includes new units)
- 2-3% = 600 to 900 units are moving to 'critical' each year based on age and location.
When one considers that approximately 6,000 GSUs exist and their design life is between 28 to 30 years, and that many in this age range are associated with low cost generation assets such as Nuclear, Hydro, and Fossel fuel stations, one begins to understand the risk profile from an economic perspective.
From a homeland security point of view, GSU and transmission transformers represent some of the most critical asset types associated with system security and reliability. When taking into consideration their replacement cost and lead time, the economic impact of terrorist acts against these asset types could be catastrophic.
Transformer risk mitigation solutions
There has been expressed interest in the industry of sparing transformers either on an individual utility basis or on a pooled basis to be shared between utilities. This approach has merit, but would require or assume harmonization of ratings and design standards for both GSU and other transmission applications. This may or may not be practical depending upon the original system constraints used for the transformer design. To make such a proposal viable, it would make sense to develop a process where the transformer population could be narrowed down into design pools based on condition and importance to system reliability. The results from this process could be used to drive the creation of common spare unit designs. However, selection of spare units may not be the most cost effective solution given that each company has its own asset management objectives and risk adversity criteria. In other words, a spare unit strategy for one company may not make sense for another company. A process whereby a company can rank assets by contribution to network reliability, condition, and fit into a company's asset management objectives is what is really required.
Hence, it is recommended to first identify the transformer asset risk profile for each utility, and then select the most appropriate asset management strategy, understanding which assets are most important.
By analyzing a specific asset type, such as transformers, the data acquisition and integration effort referenced earlier becomes much more focused and can be done in a more pragmatic and timely manner to deliver actionable recommendations that support prioritization (see example below).
Siemens has introduced an analysis technique called Asset Risk Mitigation Analysis (ARMA). ARMA has customers focus their data gathering and analysis efforts on specific assets (transformers being the first) to quickly prioritize the appropriate asset management strategy. This strategy is based on the customer's risk profile and company objectives associated with the asset type.
To find out about how ARMA Services may give you a competitive advantage, please contact your local Siemens sales representatives or contact Siemens Service Solutions at 1-800-333-7421.
What are some of the financial criteria associated with understanding risk profile and asset prioritization?
With respect to transformers, the following can be used generically.
- Assume a spare GSU costs approximately $2.0 million and the internal cost of capital is 6%. The annual interest expense of tying up capital on a spare unit would be .06*$2.0= $120,000 per year. Minimizing the units to be spared can have a significant impact on ROA and ROI.
- Assume the following potential sources of electrical power and related marginal costs of energy production (fuel only);
a. Nuclear @ $4.68 MWhr
b. Fossil Steam @ $17.35 MWhr
c. Gas Turbine @ $43.91 MWhr
d. Open Market Purchase Average @ $87.00 MWhr
Assume a GSU fails (400 Mya Unit replacement or spare unit cost $2.8M) and requires 10 days to replace with a spare unit, further assume the power must be purchased from an alternate generation source. The following costs would result from use of a Gas Turbine generation plant to replace the energy:
($43.91-$4.68)*(400*.97)*10*24 = $3.65M
Removal and Installation:
Remove old unit = .40M
Install spare unit = .40M
Total Cost = $4.45 M
Using similar calculations the following table would result.
Assume further that the GSU's for the nuclear plant are at or exceed their design life and therefore have a very high probability of failing.
If the asset manager was considering the potential purchase of another spare unit which could reduce the outage time from 10 to 5 days, the decision would be dependent upon the alternative energy source that would be used.
Continuing the example; if the fossil system was the alternative source, the reduction of 5 days would reduce the replacement energy cost from 1.18 to .59M. If the alternative energy source was gas turbine the replacement energy saving would be half of $3.65M or $1.83M, and finally if the alternative energy source was open market purchase, the savings would be half of $7.67M or $3.83M.
Hence, in the above example the purchase of a second spare would make the most sense if the alternative energy source was the open market, since replacement energy costs would exceed the cost of purchasing an additional spare transformer ($3.8M vs. $2.8M).
In actuality, the calculation would involve the probability of failure, but this was omitted to show the importance of performing a network reliability analysis (load flow and probability of failure impact on load flow). Hence, knowing condition of the asset, although important, is insufficient to make a comprehensive decision, reliability contribution and relative cost differential of alternate energy sources is required as well.
- Data from Hartford Steam Boiler.
- DOE/EIA -0348 (2003) Electric Power Annual 2003; Table 8.2 Average Operating Expenses for Major Investor Owned Utilities 1992-2003.