The Devil in the Transmission Data
Untapped T&D measurement data could make the difference on reliability.
July 2004
By David Kreiss and J.D. Hammerly
Utility executives rely on sound decision making to determine how resources should be allocated, to ensure that systems operate with a maximum efficiency and reliability at the lowest cost. These executives walk the fine line of deciding where money should be spent to minimize the likelihood of an expensive catastrophe while also achieving a targeted level of reliability. These issues include:
- Planning for capital transmission and distribution expenditures, especially breakers, transformers, circuits, and substations;
- Projecting maintenance ex-penses; in particular, de-ciding when equipment should be repaired, upgraded, or switched out with other components; and
- Performing long-term planning for power systems that are typically operating above capacity.
While utility executives must take into account the interaction and dynamics of connected RTOs and ISOs, their decisions generally are dictated by certain underlying beliefs of the current condition of their transmission and distribution (T&D) system components and their failure rates. The problem is that these condition assessments usually are based on historical trends of failure rates and maintenance expenses that are assumed will continue in the future.
Prone to Failure
But if the past is not representative of the future, then on what basis are executive decisions made? A number of factors indicate that operations patterns in the past will not represent the future. First, T&D equipment and apparatus are aging. Much of this equipment was installed in the 1960s and 1970s and is approaching the downside of useful life. This equipment in its latter years requires more monitoring and maintenance and is more prone to failure.
Compounding the problem of aging infrastructure, utilities are cutting back on personnel, including maintenance staff. Early retirement packages are resulting in the loss of senior engineers who participated in the design and purchase of a large portion of the utility's T&D system and may have more intimate knowledge of how to diagnose and solve the less-frequently encountered problems.
Graying executives who remain in the workforce usually have higher pay rates and are prime candidates for downsizing or for incentive packages that encourage them to leave their organizations. The loss of accumulated knowledge means the organization has much less practical experience to rely upon when planning, tackling a crisis, or handling everyday decision making.
In addition, demands of the market have changed the way systems are operating. During peak demand periods and when circuit maintenance is occurring, systems are being operated at historically higher capacities that provide reduced margins for error before problems occur. What can't be overlooked is the fact that congestion in T&D segments is at an all-time high, and the pressure is on to operate more reliably.
The bottom line is that the models and methodologies utility executives use to manage their utility may be built on false premises. By using faulty data, utilities risk:
- Lack of maintenance on deteriorating equipment, or unnecessary maintenance;
- Periods of system instability;
- Inefficiently allocated resources;
- Illogical project-funding priorities; and
- Unnecessary construction.
Fortunately, utility executives have an option available to make better-informed decisions by tapping into the information technology (IT) infrastructure that most utilities have built but that few have exploited. This IT infrastructure is based upon the analysis and correlation of the huge quantities of untapped measurement data found in their substations. Referred to as non-operational data, this data can be used to more precisely understand the current state of their system and quantify the risk and potential reliability problems with apparatus and circuits in the future.
When combined with operational data, this non-operational information allows for ground-up analysis of the T&D system, which permits executive-level decision-making processes to be arrived at more accurately. Processed correctly, this data not only offers the answers to questions regarding the current state and health of a utility's T&D system, but it has the ability to diagnose a problem before a catastrophic event occurs.
The benefits of collecting and storing all this data become obvious when looking at overall operations and maintenance strategy. With analytical applications, a utility can examine a chronology of events related to a piece of equipment, such as its degradation and upkeep over a long period of time, where problems can mount over months or years. As long as things go well operationally, and key parameters are met, things are fine.
When things change abruptly for some reason, though, there is cause for concern-not just for short-term operations, but for long-term maintenance, reliability, and investment strategies. Is the equipment degrading? Will it fail? Is it time to replace it, and if so, with what?
Having an archive of non-operational data allows a utility to look at the use of a piece of equipment, like a transformer, over time. That allows decisions to be made on when to repair or replace a particular transformer, and it allows analysis of whether to move a typically underused transformer already in service to a higher-demand location, or to buy a similar or better unit to meet location requirements. If the underused transformer moved from within the system must be replaced, is purchasing a lower-rated unit, at lower cost, an option for that location?
In the past, this type of analysis would have required an engineer to travel to a substation or multiple locations to retrieve data from various sources, and then analyze the data and produce a report. Today, technology allows this type of task to be done in a wholly automated manner.
Relating to Sarbanes-Oxley
This solution can play a positive role in another factor that now impacts many executives' decision making: Sarbanes-Oxley. Intended to deal only with financial data, the law also affects electric utilities and how they monitor, record, and make use of operational data. Sarbanes-Oxley has made the CFO and CEO of an organization personally exposed when the company announces financial information.
The question then becomes, what is financial information? In today's deregulated environment, line capacities in a grid have a direct tie to, and impact on, an electric utility's financial health. The argument is then fairly clear that information regarding the electric system is covered in some way by Sarbanes-Oxley.
Sarbanes-Oxley specifies that appropriate controls be in place regarding the governance of a business, thus ensuring that checks and balances exist so that no one person can subvert the process and commit fraud. The whole scenario can serve to make executives nervous and cautious. If the CEO or CFO of a utility is required to certify that the information being provided to investors and the public is accurate, how do they ensure that it is, in fact, accurate?
By collecting non-operational and operational data from substations, organizations have tangible proof of the accuracy of the information they are providing and using to make key decisions.
Analysis and Correlation of Untapped Data
Today's T&D system has huge quantities of untapped measurement data that is not collected by the utility's SCADA or EMS system. These systems generally address only operational data that is restricted to the real-time analog values and status events, while the more complex waveform and fault data are excluded from analysis.
Non-operational substation data is found in most intelligent electronic devices (IEDs) including power monitors, power quality monitors, equipment monitors, and especially common protective relays, which are able to capture fault waveform records and detailed historical logs of most every power parameter. These IEDs are capable of connecting to apparatus contacts and analog sensors to allow for fairly sophisticated equipment-specific diagnostics. In addition, today's apparatus often comes with standard or optional monitoring devices and sensors, including state-of-the-art dissolved-gas analyzers that can provide new insights into the internal problems in transformers.
With so much to gain from this non-operational data, why aren't utilities taking advantage of it? Why aren't utility executives using it in their decision-making process? Put simply, there has been no unified system architecture to overcome the four primary barriers of data access, integration, storage, and analysis-until now. Fortunately, new technology in the areas of protocol standardization, file archiving, high-speed data transmission, and automated analysis make this data not only accessible, but also much easier to integrate and process.
Non-operational substation data has been left idle partly due to the huge quantities and the complex nature of the data that made it impossible to analyze manually, except for the analysis of a specific event. However, the technology is available and has been implemented to embed expert reasoning to automatically analyze and correlate this data.
Specifically, this solution provides insights to operations to help system restoration and alarms in other departments to identify and predict system problems.
The engine allows for the configuration of an analysis module to a utility's specific T&D system and operational procedures. It is essentially the linking of the reusable intelligent objects through a "script" that allows for a high level of customization without the time and expense of traditional software development.
Ground-Up Approach
Utility executive who make decisions need information at a high level. The untapped non-operational data discussed is too raw to be of use at the executive level. What executives need is a processing of this raw data to a degree where they can derive informed decisions from it. This is not a simple task because it requires data to be processed at several levels into meaningful information, which can involve several levels of analysis. The new technology designed to exploit non-operational data takes this multi-level analysis into account when serving up information.
The first level includes the analysis that addresses a specific problem or requirement. For transformers this could be diagnosing a bushing problem, or a tap problem, or an internal winding short. The next level would be an "aggregation" of results from the first level analysis. For transformers, this could be an overall predictive maintenance module or one that determined at what capacity the transformer could be safely operated. This second level is more complex and could not be done without aggregating the lower-level analysis elements.
The final level is the executive level that would be an aggregation of a selected group of mid-level modules. Such modules would address the critical business interests of the utility to include:
- Management of system stability
A near-time view of system stability, how it is changing, and actions and resources available to improve stability if needed. Executives will be aware of deteriorating system reliability and whether there are opportunities that are available to avoid an outage or catastrophic failure. Opportunities include available energy to be purchased or out-of-service transmission assets that can be restored.
- Maintenance cost optimization
A hierarchy of what equipment requires maintenance based upon the impact of its failure and cost or maintenance.
- Reliability improvement optimization
A roadmap of system improvements based upon projected reliability and congestion problems.
- System performance optimization
Executive level interface describing the operation in terms of capacity and opportunities of improving system performance.
If implemented properly, this new technology and method of processing data will provide utility executives with intelligent supportable information on which strategic operations decision can be made.
Act Now?
In general, electric T&D systems have been designed well and perform adequately, but significant portions of systems across the nation are growing old. As operators are forced to deal with increasingly less comfortable margins of reliability because of the growing need to run systems closer to the edge, it becomes necessary to make choices about acceptable levels of risk. Unless that risk can be identified and quantified through the use of operational and non-operational data, there is no solid basis on which to make those types of risk-management decisions.
The use of non-operational data also becomes strategically important as input to the executive decision-making processes. Since the data is already available, the cost of implementing a system to analyze and correlate the data should not be significant, and in fact can have a defined ROI just based upon addressing tactical solutions.
A solution that provides for this use of both operational and non-operational data from the grid is now available in a commercial, off-the-shelf form that has been installed and is running at a number of sites. Current market conditions make it clear that the electric industry has a mandate to apply technology such as this as quickly as possible to aid in the effort to proactively work to prevent blackouts and brownouts, and provide a more reliable, stable supply of electricity to the nation.
While it took some time for utilities to generally embrace SCADA and EMS systems to make productive use of operational data, the industry may not have the time to tap the exponential value of advanced substation data at its leisure. Infrastructure is aging faster than most utilities can keep pace, and this means they must move quickly toward implementing technology that can leverage non-operational data.
David Kreiss is president of Kreiss Johnson Technologies, a developer of substation data acquisition and analysis software for the utility industry based in San Diego, Calif. Kreiss can be reached at dkreiss@kjt.com or 858-535-2088. J.D. Hammerly is regional vice president for Areva T&D, which is headquartered in Paris. Hammerly works out of the company's Seattle office. Hammerly can be reached at JD.Hammerly@areva-TD.com or 425-739-3600.
Blackout Feedback Falls Short
The Final Report on the Aug. 14, 2003, blackout and associated documents published over the past months, including FERC's Policy Statement on Matters Related to Bulk Power System Reliability (April 19, 2004), reinforce the notion that utilities are not sufficiently aware of the state and weaknesses of their T&D system and that system monitoring and management need to be improved. The first two causes of the blackout are listed as inadequate system understanding and inadequate situational awareness. As long as utilities rely solely on operational data and ignore the more advanced substation data, there will be a ceiling on their understanding of system conditions and weaknesses. With today's higher speed communications and intelligent software, much of this more advanced non-operational data can be processed fast enough to support real time operational decisions, which can enhance system awareness and understanding.
The report suggests that the described "inadequacies" seem to build over time and could have been identified and addressed through continuous aggressive system monitoring and analysis. According to the blackout report, "Although the causes discussed below produced the failures and events of August 14, they did not leap into being that day. Instead as the following chapters explain, they reflect long-standing institutional failures and weaknesses that need to be understood and corrected in order to maintain reliability." This lack of understanding could have been avoided as well. Tapping into vast quantities of non-operational data now available in the T&D system will not only provide a much clearer picture of the state of the system but will identify and address system weaknesses, including the prediction of apparatus failures and the identification of improper relay settings, before they contribute to a catastrophic blackout.
Vegetation-related faults were a key contributor to the blackout. In response, the Utility Vegetation Management Final Report-March 2004 "strongly recommends that the industry 'average' or standard needs to be substantially improved." Vegetation management cannot be expected to avoid 100 percent of vegetation-related faults, so it behooves utilities to monitor circuits and determine acceptable levels and to be able to identify a trend that suggests additional VM is required. Again, what is needed is an automated system that would establish an index or level at which vegetation-induced faults can no longer be tolerated by the system. Such an index would monitor ongoing trends and trigger an alarm when the threshold has been exceeded.
FERC is aware that utilities will need additional incentives to implement and fund new monitoring and management systems. FERC's policy statement describes its "support (of) NERC to translate the existing reliability standards into clear enforceable standards" as well as to ensure cost recovery of prudent reliability standards to improve grid monitoring and management.
It is clearly stated that the commission "assures public utilities that they will approve applications to recover prudently incurred costs necessary to further safeguard the reliability and security of our energy supply infrastructure in response to the heightened state of alert." This gives utilities the incentive and funding to improve their overall T&D information and management system. However, without the ability to collect, analyze, and correlate the more advanced non-operational substation data, these systems will fall short of meeting their goals to improve reliability and reduce costs. -D.K. & J.D.H.
Articles found on this page are available to subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.
Public
Utilities Fortnightly & Spark
